Journal of South American Earth Sciences 100 (2020) 102591

Contents lists available at ScienceDirect

Journal of South American Earth Sciences

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

Recharge sources and hydraulic communication of karst aquifer, São Miguel T

watershed, MG, Brazil
Thiago Nogueira Lucona,∗, Adivane Terezinha Costaa, Paulo Galvãob,
Mariangela Garcia Praça Leitea, Thiago Madeiraa, Leonardo Brandão Nogueiraa
a
Federal University of Ouro Preto, Department of Geology, Morro do Cruzeiro Campus, Ouro Preto, MG 35400-000, Brazil
b
Federal University of Minas Gerais, Institute of Geosciences, Pres. Antônio Carlos St. 6627, Belo Horizonte, MG 31270-901, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: The main public water supply in the São Miguel watershed (MG, Brazil) comes from a well-developed karst
City of pains region, where information about hydraulic communications between caves entrances, sinkholes, recharge and
Karst aquifer discharge areas are still unknown. The study aims to identify regional hydrogeological processes describing
Stable isotopes groundwater hydrochemical evolution, interactions between meteoric, surface and ground waters, recharge and
Geophysics
discharge sources, and possible hydraulic communications for proper management of water resources.
Brazil
Geophysical data were analyzed and water samples (meteoric, surface, and groundwater) throughout the wa-
tershed were collected for stable isotopes (18O and 2H) and major ions analyzes. Results indicated strong in-
fluences of regional geological structures in the karst water flow directions; high concentrations of major ions in
the central region of the watershed are common; karstification processes and aquifer recharge are more active
during the rainy season, while in the dry season the watershed is supplied by groundwater.

1. Introduction within the aquifer making it difficult to understand the water circula-
tion (Kresic, 2013; Han et al., 2014). Since the karst network presents
In karst environments, surface water can rapidly connect to high permeability in the epikarst zone (most weathered), recharge is
groundwater, making karst aquifer very susceptible to contamination much faster than in porous and/or fractured aquifers (Ford and
(Ryan and Meiman, 1996) where enlarged fractures and/or conduits Williams, 2007), which can correspond to velocities up to 230 m/h in
are the main entry points for contaminants (Goldscheider and Drew, Sicily, Italy (Ruggieri et al., 2017), and 1055 m/h in south China (Luo
2007). Karstification can result in different transmission and storage et al., 2016).
conditions for groundwater and can be classified according to the type Regarding hydrological processes in karst environments, study of
of recharge based on the origin of the fluids. Considering an epigenetic environmental isotopes in water (δ2H and δ18O) associated with hy-
karst system, which dissolutions take place near the phreatic surface via drochemistry have been successful in indicating the contribution of
recharge from meteoric water, if only the karst area is recharging by surface water to groundwater (Gibson et al., 2005; Feng et al., 2012;
rainwater, it is called autogenic, but if the recharge occurs also from Wong et al., 2013; Souza et al., 2015; Galvão et al., 2017; Filippini
non-soluble adjacent rocks to the karst aquifer, it is called allogenic (for et al., 2018). The variation of δ2H and δ18O in the water are in-
more details, see Palmer, 1991; Klimchouk and Ford, 2000; White, corporated in geologic and biologic systems in a predictable manner
2003; Ford and Williams, 2007; Goldscheider and Drew, 2007). (Killingley and Newman, 1982; Roden et al., 2000; Ehleringer et al.,
The conduit network responsible for recharging and storing 2008a), which allows their use as natural tracers in hydrology, arche-
groundwater in karst regions is extremely heterogeneous and aniso- ology, ecological studies, and forensic investigations (Ehleringer et al.,
tropic, being hard to describe and quantify flow features with tradi- 2008b). With the use of environmental isotopes, it is possible to track
tional methods. In addition, these heterogeneities and anisotropies, the origin and fate of sources of atmospheric humidity, subterranean
which due to the dissolution of soluble rocks generating karst cavities/ flows, identify recharge and discharge areas, and even the reliance of
networks, are often unpredictable. The interaction between water and vegetation on groundwater (Harvey and Sibray, 2001; Yepez et al.,
rock is dynamic and changes according to the water residence time 2003; Burnett et al., 2004; Williams, 2004; Bowen et al., 2012).

∗
Corresponding author.
E-mail address: [email protected] (T.N. Lucon).

https://doi.org/10.1016/j.jsames.2020.102591
Received 11 November 2019; Received in revised form 16 March 2020; Accepted 29 March 2020
Available online 05 April 2020
0895-9811/ © 2020 Elsevier Ltd. All rights reserved.
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Another tool used in karst areas is the geophysical methods The course of the São Miguel river drains perennially on the surface,
(gravimetric, electrical, and geomagnetic) for geologic modeling, mostly (over nearly 53 km) following a general S/N direction, with
identifying subsurface voids and their extensions (Chalikakis et al., meandering morphology (Menegasse et al., 2002; Dias and Velásques,
2011; Kaufmann et al., 2011; Kaufmann, 2014). The association be- 2002). The river has an altimetric range of 245 m, beginning at the river
tween geophysics and environmental isotopes becomes an excellent source at 875 m until emptying in the São Francisco river at 630 m. The
tool for the management of water resources and land uses, especially in main tributaries in the western margin, following from upstream to
karst aquifers. Geophysical anomalies indirectly reflect the depth and downstream are Moendas creek, Mina stream, Mandengo stream, and
possible runoff pathways for groundwater, while isotopic signatures Fundão stream. In turn, on the eastern margin, the main tributaries are
coupled with dye traces, geochemistry, structural, and speleological Barra stream and Candongas river, which encompass the sub-basin of
analyses may confirm the presence and interconnection of conduits and the Santo Antônio stream, both located mostly within the municipality
potential flow routes (Auler, 1994; Carvalho et al., 2015; Penãranda, of Arcos.
2016; Somaratne and Mann, 2016; Lucon et al., 2018). The identifica- Mean annual rainfall, based on a historical series (1975–1989), is
tion of various recharge sources and interactions between surface water 1.325 mm. The rainy season occurs from October to March, with mean
and groundwater is extremely important for the adequate management precipitation of 1.195 mm, corresponding to 84.98% of the annual
of water resources. Several researchers are currently concentrating ef- rainfall. The lowest precipitations occur from April to September, with
forts to understand processes that take place in karst environments, mean value of 34.5 mm, equivalent to 15.02% of the annual total. In
such as hydrochemistry, geomorphologic structure analysis, hydrologic turn, mean annual temperature is 21.28 °C, with June presenting the
dynamics of the karst flow, and estimates of mean rate and spatial lowest value (17.4 °C) and January, the highest (24 °C). These data are
distribution of recharge areas (Liu et al., 2009; Li et al., 2010; Zhao similar to those of Belo Horizonte (Fig. 2). The water balance of the São
et al., 2010; Somaratne and Mann, 2016; Galvão et al., 2018). Miguel river varies monthly with the input from rainfall, clearly
The groundwater pollution risk is considered as the interaction of marking two distinct periods: water surplus (November to April) and
aquifer intrinsic vulnerability and the contaminant load that is/will be/ water deficiency (May to September), with replenishment in the month
can be applied to the surface as a result of human activity. The con- of October (Menegasse et al., 2002).
taminant load can be controlled or modified, but the vulnerability of Geologically, the area is located at the southwestern portion of the
the aquifer is essentially fixed according to its natural hydrogeological intracratonic São Francisco basin, where carbonate rocks and fine to
setting (Foster et al., 2002). Thus, an understanding of dynamics and coarse-grained siliclatics of the Ediacaran Bambuí Group are exposed
hydraulic connections, especially in karst regions, can help prevent and unconformably overlie Archean/Paleoproterozoic basement units
problems related to the contamination of the water resource and can (Ribeiro et al., 2008; Heilbron et al., 2017) (Fig. 3). In this portion of
also help in the management of recharge and groundwater flow areas. the basin, the Ediacaran sedimentary successions are affected by NS-
The São Miguel watershed, upper São Francisco basin, center-wes- trending reverse faults and associated folds, within the east-directed
tern portion of the state of Minas Gerais, Brazil, presents a highly de- and thin-skinned Brasília foreland fold-thrust belt. The NS-trending
veloped karst network with extensive limestone massifs with karren and structures are affected by late strike-slip systems along kilometers-scale
ruinform structures, springs, sinkholes, intermittent watercourses, sinks and NW-striking corridors, which are connected with the lower Ar-
and resurgences, dissolution dolines, uvalas, poljes, and various caves chean/Paleoproterozoic basement (e.g., Muzzi-Magalhães, 1989; Reis
(Saadi, 1991; SEE, 2012; Pereira et al., 2019). These are frequently et al., 2017).
integrated forming highly permeable conduit networks, making it dif- The Doresópolis-Pains System corresponds to a left-lateral strike-slip
ficult to understand the local water circulation. Previous studies have zone that affects the Bambuí Group and older Precambrian rocks in the
indicated occasional changes in the watershed regarding thermo- Southern São Francisco Craton (Muzzi-Magalhães, 1989). Underlying
tolerant coliforms, high concentrations of total phosphorus, nitrate, the São Miguel watershed, this system extends toward the Pains town,
arsenic, manganese, cadmium, aluminum and iron, turbidity, decreased where its lateral displacement reaches up to 4 km and juxtaposes highly
levels of dissolved oxygen, and microbiological contamination pro- deformed limestones and undeformed sedimentary rocks. In this area,
moting cases of schistosomiasis (Haddad, 2007; Haddad and Magalhães its surface expression is defined by linear hills, ridges, and drainages
Jr. 2010; Lucon et al., 2018). Despite all this information, there are still (Fig. 3).
gaps regarding confirmations of the preferential routes of these con- The São Miguel watershed is divided into four aquifer systems
taminants in the local karst network, from the recharge areas to the (Mourão et al., 2001): 1) granular or porous, 2) karst, 3) karst-frac-
discharge areas. tured, and 4) fractured. The karst and karst-fractured systems consist of
The study sought to broaden the understanding of hydrological Neoproterozoic limestones divided into two members (Ribeiro et al.,
behaviors in karst regions and identify hydrogeological processes that 2003): Pedro Leopoldo (at the base), composed of light-colored calci-
take place in the São Miguel watershed. To do so, the objectives of the siltites, marls, and pelites. In this member, lenses of silexite interleaved
work are to understand possible origins, groundwater flows, hydro- with pelite were reported by Carvalho et al. (2015); and Lagoa Santa (at
chemical evolution mechanisms of these waters, and the interactions of the top), characterized by medium to fine dark limestones and dolo-
meteoric and surface waters with groundwater and epigenic recharge mites. The granular or porous aquifer system is represented by detrital
sources, with the aid of stable isotopes 18O and 2H, major ion con- covers composed of alluvial and colluvial eluvial sediments that are
centrations, and geophysical analyses. As a result, the study contributes found along the water courses, in floodplains and terraces. The frac-
to new interpretations helping improving water resource management tured aquifer system, inserted within the granite-gneiss complex, out-
tools and land use and occupation planning. crops in the extreme eastern portion of the São Miguel watershed
(Fig. 3).
2. Site description Geomorphologically, the area is within the Arcos-Pains
Speleological Unit, which is divided into six morphostructural units
The study area comprises the karst area of the upper São Francisco (Fig. 4): 1) Lagoa da Prata tabular plateau, of flat and low surface with
basin, specifically the basin of the São Miguel river, in the center- rare limestone outcrops, long floodplains with low density of water-
western region of the state of Minas Gerais, Brazil (Fig. 1), 180 km courses and presence of various lakes; 2) Doresópolis-Iguatama fluvial-
northwestwards from the state capital Belo Horizonte. Most of the basin karst plateau, of flat to smooth relief dominated by calcipelites and
is located within the municipalities of Pains, whose public water supply presence of karst outcrops, drainage sinks and natural lakes; 3) and 4)
is managed by the Water Supply and Sewage Service [Serviço de Abas- Western and Eastern São Miguel karst plateaus, both from the Arcos-
tecimento de Água e Esgoto, SAAE]. Pains-Doresópolis Karst Province, of karst relief associated with water

2
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 1. Location in the UTM coordinates of the upper São Francisco basin (upper-right smaller maps), and of the study area in the São Miguel watershed in the central
west of the state of Minas Gerais (main map). The main map also shows the municipalities and sampling points for groundwater (black, white and yellow triangles),
surface water (blue circles), and meteoric water (white squares). (For interpretation of the references to color in this figure legend, the reader is referred to the Web
version of this article.)

action and geologic discontinuities of the substrate. The São Miguel exposed soil, limestone outcrops and flooded areas (Pereira et al.,
river separates these two units of homogeneous morphology; 5) Arcos- 2019). Understanding the land use and occupation patterns of the basin
Córrego Fundo Dissected Plateau, associated with the basement, helps in characterizing the various sources of water pollution, where
strongly carved by valleys characterized by flat and convex tops and pastures are characterized as diffuse pollution. Mining, in turn, as a
low drainage density; and 6) Floodplains, with a thick layer of soil, non-point source due to the exploitation of rocks and calcination, can
located on the riverbeds of the main rivers (Pizarro, 1998; Martins, promote atmospheric pollution, which can later pollute soils and water
2013). bodies. In the case of urban areas, these are considered as main point
Speleologically, the Western and Eastern São Miguel karst plateaus, sources of pollution due to the discharge of sewage, and urban and
located in the Lagoa Santa Member, is the richest region in speleolo- industrial effluents. Finally, vegetation represents areas that are not
gical terms, with more than 700 registered cavities (Figs. 3 and 4). characterized as sources of pollution (Haddad, 2007; Haddad and
Cavities developed in bedding planes of subhorizontalized rocks, which Magalhães; 2010).
are eventually fractured, are more common. It is in that region that
most of the sinkholes and perennial springs with active water flows are
concentrated. In the remaining domains, coincident with the Pedro 3. Materials and methods
Leopoldo Member, more pelitic, the occurrences of sinkholes or dolines
are smaller (Teixeira et al., 2013). To understand the hydrodynamic behavior of the karst region of the
Regarding the use and occupation, pastures (61%) are predominant, São Miguel watershed, and the site's hydrogeologic setting and inter-
followed by vegetation (27%); mining and urbanization together ac- actions, a total of 17 points were monitored during one rainy season
count for nearly 3.5% of the basin, while the remaining 9.5% consists of (March 2016) and one dry season (July 2016). Three samples of me-
teoric (rainfall) water were taken, two fluvial samples, and 12

3
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 2. Temperatures (lines) and precipitations (col-

umns) in the cities of Pains (red) and Belo Horizonte
(blue). Belo Horizonte is located about 180 km
Midwest from Pains. The data show similar climate
conditions. (For interpretation of the references to
color in this figure legend, the reader is referred to
the Web version of this article.)

groundwater samples (1 dug well, 3 tubular wells, and 8 caves/sink- 3.1. Stable isotopes
holes), totaling 34 samples. Physicochemical parameters were mea-
sured for all samples in situ and with later chemical analyses to obtain Samples containing 30 ml of water were collected and stored in
the values of major ions and isotopes 18O and 2H. The techniques used amber glass flasks closed with a stopper and cap. The main concern was
for the hydrochemical, isotopic, and geophysical analyses are described to avoid isotopic fractioning. To do so, these flasks were filled with the
below. sampled water, avoiding air bubbles within.
Meteoric water was sampled by installing rain collectors. These

Fig. 3. Geological and structural contexts in the São Miguel watershed. Lineaments with NW-SE directions are more common, while the São Miguel river follows a
main N/S direction.

4
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 4. Geomorphological units map found in the São Miguel watershed.

devices consisted of 5-L gallons were placed within polystyrene boxes the first meteoric line proposed by Craig (1961) through the equation
buried in the ground in order to preserve the temperature and avoid δ2H = 8.17 δ18O + 10. In turn, the BHMWL was obtained from the
sunlight. These gallons were connected to a polypropylene funnel GNIP database (2008–2017) and presents the following equation:
through silicone hoses measuring 6 m in length. The funnels were δ2H = 8.03 δ18O + 12.67.
covered by a net to avoid that leaves, twigs, insects and other items fell
inside the gallons. According to GNIP (Global Network of Isotopes in
Precipitation), the 6 m of the hose used made evaporation and the loss 3.2. Groundwater sampling for major ions
of lighter isotopes insignificant (IAEA/WHO, 2004).
Isotope analyses (18O e 2H) were carried out at the Geochemistry The concentrations of nitrates, sulfates and chlorides were recorded
Laboratory of the Federal University of Ouro Preto – Brazil. A Thermo with a colorimeter (DR 890 HACH). To determine the concentration of
Finningan Delta V advance Mass Spectrometer (IRMS) coupled to major elements (Ca2+, Mg2+, K+, Na+), the samples were pre-filtered
GasBench II, processed by Laboratory Information Management System with a 0.45-μm membrane and kept in 30-ml polyethylene flasks,
(LIMS) and calibrated in relation to the Vienna Standard Mean Ocean acidified with nitric acid ultrapure grade (HNO3− – pH < 2) to sta-
Water (VSMOW) was used. The results were expressed in notation δ in bilize dissolved metals and, later, stored at 4 °C (CETESB, 2011). The
parts per thousand (‰) in relation to V-SMOW. The fractionation factor concentration of HCO3− was determined in the field using titration
used was 0.9981 for δ 18O and 2.7365 for δ 2H to correct the raw data. (Greenberg et al., 1995).
Hydrogen and oxygen isotopes of water samples results were expressed All samples were analyzed through inductively coupled plasma
as δ18O and δ2H, where δsample (‰) = ((Rsample/ atomic emission spectroscopy (ICP-OES Agilent725), at the
Rstandard) − 1) × 1000, where R is D/H, 18O/16O. The external accu- Geochemistry Laboratory of the Geology Department at the Federal
racy based on multi-standard measurements of V-SMOW (Vienna University of Ouro Preto. After conducting all measurements, the pre-
Standard Mean Ocean Water), GISP (Greenland Ice Sheet Precipitation) cision of the samples was verified by calculating the ionic balance
and SLAP-2 (Standard Light Antarctic Precipitation) was better (Custodio and Llamas, 2001). A good analysis is expected to yield a
than ± 0.1‰ for δ18O and ± 1‰ for δ2H. V- SMOW, GISP, and SLAP-2 balance with differences below 2% (Bittencourt, 1994). Schoeller
are water reference standards prepared by the International Atomic (1962) and Weight and Sonderegger (2000) affirm ionic balance error
Energy Agency (IAEA) defined to control the successful calibration in cannot be above 5% and should never exceed 10% in order to consider
terms of precision and accuracy to be followed by the laboratories an analysis analytically correct (Feitosa et al., 2008). However, for
worldwide. poorly mineralized waters, in which it is difficult to quantify ions with
After the analyses, samples were compared with the global meteoric low concentrations, errors of the order of 30% are acceptable (Custodio
water line (GMWL - Rozansky, 1993), available at the IAEA website: and Llamas, 2001).
http://isohis.iaea.org, and with the Belo Horizonte meteoric water line All cation and anion results were treated and transformed from mg/
(BHMWL). A meteoric water line was not established for the study area. L to mEq/L to elaborate the hydrogeochemical model. Stiff diagrams
The BHMWL was chosen given that it presents similar climatic condi- were generated using QualiGraf V.1.17 (Mobus, 2003), to better illus-
tions (Fig. 2) and altitude (Pains ≈ 740 m. a.s.l. and Belo Horizonte trate the hydrochemistry of the region studied.
≈820 m. a.s.l.). The meteoric line established by Rozansky et al. (1993)
is linearly related by the equation δ2H = 8.17 δ18O + 11.27, updating

5
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

3.3. Structural geophysics 4.1. Aeromagnetic data

For geophysical and structural analysis, the Shuttle Radar The magnetic analytic signal (AS) map (Fig. 5A) shows large
Topographic Mission (SRTM) provided 3D models with quality and lineaments of NW-SE direction, obliterated by others from an approx-
spatial resolution of 1 arc-s (~30 m) and 3 arc-s (~90 m) with hor- imate N–S direction. There are also W-E lineaments at smaller scales,
izontal datum WGS84 and vertical datum WGS84/EGM96 with relative though with higher frequency, observable in the stereogram of Fig. 3A,
vertical accuracy in the order of 5 m (Smith and Sandwell, 2003). These elaborated based on 494 traced lineaments.
datasets were developed from the original spatial resolution by kriging The magnetic second vertical derivative map (Dzr2), after regional
and then geomorphometric algorithms analyses were applied on the trend removal (Fig. 5B), displays lineaments from WNW to NNW with
refined data to calculate the variables such as slope, orientation of contrasts of high (warm colors) and low (cold colors) intensity. The
slope, horizontal curvature and vertical curvature (Valeriano and anomalies demonstrate a gradual change in the SW region towards NE.
Rossetti, 2008). The Topodata Project (http://www.dsr.inpe.br/ Higher intensity contrasts occur sectioning the NW-SE lineaments. The
topodata/) delivers the Digital Elevation Model (DEM) and its geo- gradual change was interpreted as a variation in the depth of the top of
morphometric variables derived from SRTM for the entire national the basement rocks, which is deeper in the SW portion and shallower in
territory. Before downloading these data, the SRTM model was im- the NE portion.
ported in ArcGis 10.5 for integration and interpretation. The objective While AS map presents deeper structures related to the crystalline
was delineating lineaments related to geological structures. basement, Dzr2 map offer shallower structures. The first map is the sum
The magnetic and radiometric data has part of the Airborne of subsurface magnetic sources, and the second map high frequencies of
Geophysical Survey Program (2005/2006) from Minas Gerais low depth structures. Both AS and Dzr2 maps present similar directions
Development Company (CODEMIG) following the Mines and Energy of regional lineaments, the stereograms approve these directions. The
Ministry (MME), Mining and Mineral Transformation (SGM) and the caves are connected manly by NW-SE and N–S lineaments.
Brazilian Geological Survey (CPRM). These data cover an area of The A-A′ section presents a preferred NW/SE direction, encom-
69,127 km2 where 123 districts of the Southwest of Minas Gerais are passing Narigudo, Loca Feia and Mastodonte caves and São Miguel river
located, which includes the 5 municipalities that comprise the region of (Fig. 1). This corroborates the study made by Carvalho et al. (2015), in
the São Miguel watershed (Formiga, Pains, Iguatama, Córrego Fundo, which the authors presented a map of the Narigudo cave and reported
and Arcos). The survey flight occurred 100 m above the ground in a that the river conduit over approximately 830 m is controlled by frac-
constant barometric height, along parallel survey lines spacing 400 m tures with a preferred NW-SE direction. The B–B’ section presents a N/S
and perpendicular tie line spacing of 20 times the survey lines spacing. preferred direction, parallel to the São Miguel river, encompassing the
Height resolution and precision information were captured in a sample region of the Sócrates sinkhole, Éden cave, Distrito da Mina neighbor-
interval of 3.9 m for magnetic data and 7.8 m for radiometric data. hood, and Angolinhas spring. According to data available from the
The magnetic maps were used to detect magnetic susceptibility due National Register of the Caves of Brazil (CNC), the Éden cave measures
to both hydrothermal zones related to shear zones and lithological 3350 m in its mapped portion. The Eden cave is conditioned by normal
contacts and dike bodies. The main objective of radiometric maps was failures and fractures (Pires and Rudnitzki, 2019).
detecting lithology and hydrothermalized structures. The database However, according to information obtained from the Saldanha
correction and processing were provided by CPRM. Using a commercial mining company (personal communication, 2019), who financed cave
software (Geosoft, 2014), the grid of total magnetic intensity and diving studies, there is an underground river measuring over 1 km that
radiometric data were reprocessed by the minimum curvature method has not been summed to the topography above, all following a main N/
in a regular 350 m grid. The processing steps of magnetic data were S direction. Finally, the C–C’ section also presents a N/S preferred di-
performed using MAGMAP tools to apply Fourier Transform and fil- rection, encompassing the center-northern region of the basin, fol-
tering techniques routinely applied to interpret these data. The 2.5D lowing the São Miguel river until its river mouth, at the São Francisco
magnetic and radiometric maps overall were exported to the ArcGis river.
10.5 and subsequently integrated and interpreted.
With the support of the geophysical analysis and to define the main 4.2. Hydrochemistry and stable isotopes
directions of the fractures in the study basin, the stereographic pro-
jection (two-dimensional projection of the Schmidt network) and fre- Regarding the percolation process that takes place within the
quency histograms were made using Openstereo software (Grohmann Narigudo cave, with a 70 m gap, starting at the beginning of the A-A’
et al., 2011). section (Fig. 5), during both the rainy and dry seasons, the waters that
flow over the “draining” speleothem present higher concentrations of
4. Results Ca, Mg and HCO3− (Tables 1 and 2). Regarding the isotopic signatures
of 18O (Fig. 4), the least negative values were found in the speleothem
Geophysical, hydrochemical, and isotopic results were separated Ofuro, in the Narigudo cave, while the most negatives ones were in the
into three sections (Figs. 1 and 5): São Miguel river (towards its flow direction).
Still in section A-A’, during sampling, the Loca Feia cave presented
• Section A-A’: regards the percolation of waters along a 70-m gap, alterations in levels of Ca and HCO3− when compared with the beha-
initiating by the Narigudo cave (with three inside points of sam- vior of the other sampling points of the section, showing at times a
plings: the speleothems called Ofuro and Escorrimento, and an un- slight increase, and at others a slight decrease. Regarding isotopic be-
derground stream), followed by the hydraulic communication be- havior, during the rainy period, values were less negative than during
tween the caves Narigudo, Loca Feia, and Mastodonte, until the São the dry period, with the most negative values of 18O observed in the
Miguel main river; water of the Narigudo river (Fig. 6).
• Section B–B’: presents the hydraulic communication between the For section B–B’, the highest values of Ca and HCO3− were observed
Sócrates sinkhole until the Éden cave, the spring of Angolinhas, and in the SAAE well, located in the Distrito da Mina neighborhood. The
the contribution of waters from a residential neighborhood called Éden cave presented the highest concentrations of Mg (Tables 1 and 2).
“Distrito da Mina” in this system; Regarding isotopic signatures, the SAAE well presented the most ne-
• Section C–C’: covers the portion of the São Miguel river from the gative results for 18O during sampling (Fig. 7).
point where the Fundão stream joins it until the mouth of the São In turn, section C–C′ showed a gradual increase in the concentra-
Miguel watershed. tions of Mg in favor of the hydraulic gradient, with alterations/

6
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 5. - Magnetic map of AS (A) and Dzr2 (Second Vertical Derivative) with trend removed (B). Note the N–S lineaments in the center and south-west of the map. The
gradual change in background color (green to red), from SW to NE, suggests a ramping of the basement in that direction, with lower depth to NE. (For interpretation
of the references to color in this figure legend, the reader is referred to the Web version of this article.)

decreases only observed in the fishermen's ranch, located at the mouth Meteoric waters presented mean values of −3.01‰ and −13.36‰
of the watershed (Fig. 1). This same site presented a significant increase for the signatures of 18O and 2H, respectively, during the rainy season.
in relation to other points regarding concentrations of Ca (Tables 1 and During the dry season, mean values for these same signatures were
2). In this section the most negative values for both 18O and 2H were −4.03‰ and −19.23‰, respectively. These values were similar to
registered at the SAAE well – located in the region known as Sumidouro those reported by Galvão et al. (2017) for meteoric waters also from the
(Fig. 8). karst region of Sete Lagoas (18O of −2.83‰ and −11.84‰ for 2H). It is
The highest values of Ca, Mg and HCO3− in the basin occurred important to highlight that during the rainy season there was an im-
predominantly during the rainy season. Regarding isotopic analysis, all poverishment of light isotopes due to the residence time of these sam-
samples were graphically located either between the Belo Horizonte ples in the rainwater collectors in section A-A’.
and global meteoric water lines or very close to these lines (Figs. 6–8). During the percolation process, high concentrations of Ca, Mg and
HCO3− occur in the Escorrimento speleothem, in the Narigudo cave.
5. Discussion Moreover, in the deepest level of the same cave, there is a decrease in
cationic concentrations (Tables 1 and 2, Fig. 9). The increase occurs due
The discussion will be presented sequentially starting with struc- to the chemical deposits through which the water flows, while the de-
tural geology and geophysics, hydrochemistry and, lastly, stable iso- crease is due to lenses of silexite interleaved with pelite, reported by
topes, so to interpret the chemical evolution of the waters and recharge Carvalho et al. (2015).
sources based on the schematic conceptual model of the São Miguel Hydrochemical and isotopic alterations were observed among
river basin (Fig. 9). Speleogenesis, development of dissolution conduits, sampling points in section A-A’ (Fig. 9), especially in the Loca Feia cave.
and aquifer flow routes are directly associated with predominant sur- During the dry season, when there was no influence from the input of
face attitudes by means of the dominating set of fractures, as foreseen meteoric waters, a slighting impoverishment of heavier isotopes (both
18
by the model of Worthington (2005). O and 2H) was observed, indicating probably that part of the water
Regarding geophysics and structural geology, the preferred direc- comes from another place in the cave blending old and new waters.
tions occur at times over the NW-SE lineaments, at others N–S and even In section B–B’, the high concentrations of Mg within the Éden cave
W-E, similar to the preferred directions reported by Penãranda (2016), occur because of the contact of calcitic and dolomitic limestones of the
who highlighted three sets of directions (N–S, N30-40E, W-E) for the Lagoa Santa and Pedro Leopoldo members (Lucon et al., 2018). This
karst region of Sete Lagoas, within the same hydrogeologic context and section also showed high hydrochemical and isotopic similarities be-
approximately 170 km northeastwards from the study area. tween the Sócrates sinkhole, Éden cave and Angolinhas sinks and

7
 T.N. Lucon, et al.

Table 1
Sampling sites in the study area and values during the rainy season of chemical and isotopic analyses for waters in São Miguel river basin.
Rainy Season (March 2016)

Downstream Upstream Samples δ18O (‰) δ2H (‰) pH CE Cations Anions Error in % (Custodio and Llamas,
Date Ca+2 K+ Mg+2 Na+ HCO3− SO4−2 Cl− NO3− 1976)
μs/cm mg/L
Section A-A′ Rainwater 23 −3.33 −24.71 – – – – – – – – – – –
Ofuro speleothem - Narigudo cave 20 −3.56 −15.39 7.94 381.90 73.30 0.17 5.76 0.41 287.10 < LQ 0.1 2.2 8.07
Escorrimento speleothem -Narigudo cave 22 −6.07 −32.79 7.45 595.20 119.80 0.01 1.05 0.12 450.97 < LQ 0.2 4.6 9.21
Underground stream - Narigudo cave 22 −6.40 −34.19 7.43 417.90 82.55 0.23 3.82 0.63 313.32 < LQ 0.1 2.3 4.40
Loca Feia cave 20 −5.63 −31.93 7.23 414.70 81.40 0.38 4.38 0.59 301.52 < LQ 0.2 3.2 6.26
Mastodonte cave 20 −5.71 −31.30 7.36 430.20 83.83 0.38 5.27 0.57 312.00 < LQ 0.1 4.2 5.71
São Miguel river 19 −6.10 −35.75 7.51 441.30 85.84 0.53 6.72 0.89 310.69 < LQ 0.2 3.6 9.09

8
Section B–B′ Rainwater 25 −2.99 −8.04 – – – – – – – – – – –
SAAE well (Distrito da Mina) 29 −6.03 −37.06 7.09 555.80 112.53 1.25 8.94 1.42 368.70 < LQ 0.2 2.7 3.20
Sócrates sinkhole 25 −5.59 −34.22 7.73 436.80 78.80 1.45 10.28 1.16 322.49 < LQ 0.2 0.0 8.38
Éden cave 24 −5.59 −58.28 7.55 454.50 82.26 1.06 10.66 1.00 326.43 < LQ 0.3 3.7 10.41
Angolinhas spring 19 −5.61 −33.90 7.1 506.50 100.96 1.05 7.24 1.75 356.58 < LQ 0.5 3.2 5.92
Section C–C′ Rainwater 28 −2.71 −7.34 – – – – – – – – – – –
SAAE well (Sumidouro) 29 −6.87 −41.52 7.43 403.60 85.27 0.50 2.00 1.60 271.42 1 < LQ 4.9 6.79
São Miguel river after receiving Fundão 28 −5.59 −33.57 7.71 437.00 82.92 1.27 8.51 1.46 303.84 1 0.5 3.5 6.47
stream
Lagos Company well 28 −5.52 −34.83 6.99 563.60 96.18 5.21 12.26 8.81 380.71 3 0.3 2.4 2.37
Dug well fisherman's ranch 28 −6.71 −38.56 7.47 542.90 113.28 0.97 2.15 2.37 386.71 1 0.6 7.1 5.92

-: No data.
< LQ: Below limit of quantification.
Journal of South American Earth Sciences 100 (2020) 102591
 T.N. Lucon, et al.

Table 2
Sampling sites in the study area and values during the dry season of chemical and isotopic analyses for waters in São Miguel river basin.
Dry Season (July 2016)

Downstream Upstream Samples δ18O (‰) δ2H (‰) pH CE Cations Anions Error in % (Custodio and Llamas,
Date Ca+2 K+ Mg+2 Na+ HCO3− SO4−2 Cl− NO3− 1976)
μs/cm mg/L
Section A-A′ Rainwater 15 −3.68 −14.96 – – – – – – – – – – –
Ofuro speleothem - Narigudo cave – – – – – – – – – – – – – –
Escorrimento speleothem - Narigudo cave 22 −6.90 −41.60 7.97 556.00 92.46 0.30 11.60 0.26 401.73 < LQ 0.09 5.2 0.96
Underground stream - Narigudo cave 22 −7.11 −44.04 7.39 366.10 70.05 0.24 2.47 0.46 260.10 < LQ 0.21 1.1 6.41
Loca Feia cave 15 −6.78 −45.11 7.35 383.90 73.82 0.38 4.12 0.51 191.95 < LQ < LQ 4.6 1.19
Mastodonte cave 15 −6.82 −44.16 7.46 366.80 69.02 0.35 3.16 0.47 283.25 < LQ 0.06 5.2 9.01
São Miguel river 18 −6.84 −48.34 7.92 377.30 70.81 0.32 3.86 0.47 280.53 < LQ 0.04 4.4 2.87

9
Section B–B′ Rainwater 12 −4.60 −28.19 – – – – – – – – – – –
SAAE well (Distrito da Mina) 20 −7.25 −49.30 7.39 562.30 98.77 0.85 8.32 0.56 396.28 < LQ 0.02 8 1.89
Sócrates sinkhole 12 −6.98 −50.24 7.81 429.50 71.36 0.66 9.35 0.73 325.47 < LQ 0.14 5.2 5.84
Éden cave 18 −6.98 −44.82 8.04 433.60 70.42 0.72 9.31 0.72 324.11 < LQ 0.03 2.9 2.90
Angolinhas spring 14 −6.61 −44.95 7.54 537.40 93.08 0.83 7.66 1.29 379.94 < LQ 0.06 5.2 0.03
Section C–C′ Rainwater 16 −3.83 −14.55 – – – – – – – – – – –
SAAE well (Sumidouro) 19 −7.61 −52.10 7.45 398.20 75.23 0.50 1.88 1.09 290.06 < LQ 0.07 5.6 5.76
São Miguel river after receiving Fundão 16 −6.64 −45.38 7.85 432.70 73.37 0.83 8.00 1.59 337.73 < LQ 0.23 4.4 7.39
stream
Lagos Company well 19 −6.64 −46.15 7.3 479.00 87.34 2.35 11.13 3.78 321.53 1 0.18 3.1 0.17
Dug well fisherman's ranch 17 −6.25 −40.90 7.08 541.40 120.56 0.90 2.19 1.39 386.67 < LQ 0.16 7.6 9.16

-: No data.
< LQ: Below limit of quantification.
Journal of South American Earth Sciences 100 (2020) 102591
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 6. Plot of mean δ2H versus δ18O for water samples from section A-A′ in the karst aquifer in rainy (A) and dry (B) seasons, local rainwaters in comparison with the
global meteoric water line (Rozansky et al., 1993).

springs (Tables 1 and 2, Figs. 6 and 9), demonstrating a hydraulic High concentrations of Ca, Mg and HCO3 occur predominantly
communication, as reported by Cadamuro (2007) and Freitas (2009). during the rainy season, and are directly related to the dissolution of
The highest concentrations of Ca and the most negative values of 18O carbonate rocks through the karstification process of the basin
were reported in the SAAE well – Distrito da Mina neighborhood (Goldscheider and Drew, 2007).
(112.53 mg/L and −6.03‰ during the rainy season and 98.77 mg/L Galvão et al. (2017) obtained a mean 18O value of −6.45‰ and a
and −7.25‰ during the dry season), which could be an indication of mean 2H value of −42.5‰ for the groundwater of the karst regions of
deeper/older waters, since Ca is related to the dissolution of calcite and Sete Lagoas, which are close to the values found for the São Miguel
a more negative 18O value indicates a lower influence of meteoric river basin (−6.34‰ for 18O and −40.55‰ for 2H). This indicates that
waters. both karst regions may present similar hydrologic behavior, with re-
Mg in section C–C′ presented an increasing behavior during the charge directly influenced by local rainfall.
analyses, reflecting the residence time of the waters, due to its low During the rainy season, the isotope signatures of oxygen for the
dissolution rate (Mudry, 1987; Plagnes, 1997; Batiot et al., 2003; basin's waters (with the exception of meteoric waters) presented more
Galvão et al., 2017). However, Mg decreased sharply in the fishermen's negative values from upstream to downstream, with mean 18O value of
ranch (Fig. 9) at the mouth of the basin, possibly due to the influence of −5.57‰ in section A-A′, −5.70‰ in section B–B′ and −6.17‰ in
the large water volume of the São Francisco river acting as a solvent for C–C’. These values were plotted between the GMWL and the BHMWL,
this element. while section A-A′ was above the BHMWL, indicating possible modern

Fig. 7. Plot of mean δ2H versus δ18O for water samples from section B–B′ in the karst aquifer in rainy (A) and dry (B) seasons, local rainwaters in comparison with the
global meteoric water line (Rozansky et al., 1993).

10
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Fig. 8. Plot of mean δ2H versus δ18O for water samples from section C–C′ in the karst aquifers in rainy (A) and dry (B) seasons, local rainwaters in comparison with
the global meteoric water line (Rozansky et al., 1993).

local recharge (Figs. 6–8). During the dry season, groundwater and (−7.25‰), followed by the A-A′ region (−6.84‰), where the Nar-
surface water are more impoverished, meaning that they have more igudo river presented the most negative value (−7.11‰) and, finally,
negative values than during the rainy season, especially in the central C–C′ with 18O value of −6.78‰, with the SAAE well – Sumidouro
portion of the watershed (B–B′), with the most negative mean value for (−7.61‰) as the most negative of this region. During the dry season,
18
O (−6.95‰), directly influenced by the SAAE well - Distrito da Mina samples were below the GMWL, suggesting a local evaporation

Fig. 9. Hydrogeochemical model of the São Miguel river basin contemplating from the areas of recharge to the discharge in the São Francisco River. In the Narigudo
cave (section A-A′), the speleothems Ofuro and Escorrimento, and the underground river are located, in relation to the surface level, approximately 10, 40, and 70 m
deep, respectively.

11
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

signature before infiltration or even a mixture between groundwater The study design was approved by the appropriate ethics review board.
and evaporated surface water (Clark and Fritz, 1997). We have read and understood the journal's policies, and we believe that
Regarding the 2H isotope signatures, the central region of the basin neither the manuscript nor the study violates any of these. There are no
(section B–B’) presented the most negative mean values, in both the conflicts of interest to declare.
rainy (−40.86‰) and in the dry (−47.33‰) seasons. This indicates
that these waters had gone through previous evaporation (Minardi and Acknowledgments
Bontempo, 2000), not reflecting the composition of current rainfall.
Fig. 9 summarizes the hydrochemical and isotopic behavior of the Special thanks go to the Postgraduate Program in Crustal Evolution
watershed studied, in addition to about 11 km of groundwater flows: and Natural Resources of the Geology Department of UFOP, the
the high topographic levels of the basin (Narigudo cave and SAAE well Excursionist and Speleological Society (SEE), the laboratory LGqA-
Sumidouro and the Distrito da Mina neighborhood) encompass the re- DEGEO and the National Research Council (CNPQ). Conselho Nacional
charge areas (Tables 1 and 2). The waters that infiltrate the Narigudo de Desenvolvimento Científico e Tecnológico (CNPQ). Processo:
cave flow to the Loca Feia and Mastodonte caves in a NW-SE direction 141611/2014-6.
(Figs. 5, 6 and 9). In the central region of the basin (B–B′), located
parallel to the São Miguel river, these waters flow in the preferential References
N–S direction, following the regional lineament directions (Fig. 5), also
observed in C–C’. Still in B–B′, the contact between limestones of the Auler, A., 1994. Hydrogeological and Hydrochemical Characterization of the Matozinhos-
Lagoa Santa and Pedro Leopoldo members and chemical deposits Pedro Leopoldo Karst, Brazil. Msc Thesis. Faculty of the Department of Geography
and Geology, Western Kentucky University, Bowling Green, pp. 110.
(speleothems) was shown to influence hydrochemical signatures, in Batiot, C., Emblanch, C., Blavoux, B., 2003. Carbone organique total et magnésium: deux
addition to the high similarity with hydrochemical data, demonstrating traceurs complémentaires du temps de séjour dans l’aquifere karstique. Académie des
the hydraulic communication that exists there (Figs. 5, 7 and 9). Fi- Sciences, éditions scientifiques. Elsevier, pp. 205–214.
Bittencourt, A.V.L., 1994. Hidroquímica. Módulo II. In: Curso Sul-americano sobre
nally, C–C′ is the region of the basin mouth (dug well in the fishermen's Avaliação e vulnerabilidade de Aqüíferos, 1, Curitiba, 1994. Memória, Curitiba,
ranch) (Figs. 8 and 9). UFPR, ABAS, Universidad de la República Oriental del Uruguay, pp. 23–53.
Bowen, G.J., Kennedy, C.D., Henne, P.D., Zhang, T., 2012. Footprint of recycled water
sub-sidies downwind of Lake Michigan. Ecosphere 3 (6), 53.
6. Conclusion Burnett, A.W., Mullins, H.T., Patterson, W.P., 2004. Relationship between atmospheric
circulation and winter precipitation delta O-18 in central New York State. Geophys.
Res. Lett. 31, L22209.
Regional lineaments of NW/SE, N/S, and W/E directions indicate Cadamuro, L.M., 2007. Relatório de Avaliação Qualitativa de Interferências Físicas
clustering of cavity formation, as well as conditioning of important (Explotação de Água para Consumo Humano, Mineração, Irrigação) na Região da
karst flow networks, estimating about 10 km of karst hydraulic com- Área da Bacia do São Francisco. IBAMA–DIREC–CECAV. Programa das Nações Unidas
Para o Desenvolvimento – PNUD, pp. 45.
munications (Narigudo-São Miguel/Socrates-Angolinhas). This eluci- Carvalho, M.G., Meyer, B.O., Lucon, T.N., Nunes, T.R., 2015. Caracterização
dates important regions that need different actions to manage land use espeleológica do abismo do Narigudo. Pains – MG. In: Rasteiro, M.A. (Ed.), Sallun
and occupation in order to guarantee the quality of local groundwater. Filho W (orgs.) 33 Congresso Brasileiro de Espeleologia. Eldorado. Anais. Campinas:
SBE. 2015, pp. 291–298.
Karstification processes are accentuated during the rainy season in
CETESB – Companhia Ambiental do Estado de São Paulo, 2011. Guia Nacional de Coleta e
the limestones from the Lagoa Santa and Pedro Leopoldo members, Preservação de Amostras: Água, Sedimento, Comunidades Aquáticas e Efluentes
being more prominent in the first one. The conservative behavior of Mg Líquidos. São Paulo. pp. 327.
Chalikakis, K., Plagnes, V., Guerin, R., Valois, R., Bosch, F.P., 2011. Contribution of
helps to understand the residence time of the waters of the watershed,
geophysical methods to karst-system exploration: an overview. Hydrogeol. J. 19 (6),
which is enriched towards the hydraulic gradient and diluting only at 1169.
the mouth, besides helping to prove the hydraulic communications. Clark, I., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. CRC Press, New York.
Groundwater is derived from direct recharge of local precipitation Craig, H., 1961. Standard for reporting concentration of deuterium and oxygen-18 in
natural waters. Science 133, 1702–1703.
during the rainy season, as well as areas with modern recharge, such as Custodio, E., Llamas, M.R., 1976. Hidrologia Subterrânea. 2 Ediciones Omega S.A.,
the Narigudo Cave region and SAAE well -Sumidouro, as well as the Barcelona.
discharge area near the mouth of the basin. The hydraulic commu- Custodio, E., Llamas, M.R., 2001. Hidrologia subterrânea. Tomo I e II. Ediciones Omega,
S.A, Barcelona, pp. 2350.
nications of Narigudo Cave pass through the caves Loca Feia and Dias, F.S., Velásques, L.N.M., 2002. Hidrogeologia da bacia do rio São Miguel, municípios
Mastodonte emerging in the São Miguel river, as well as the Socrates de Pains e Arcos -MG. Águas Subterrâneas (1), 1–20.
through the Eden Cave receiving contributions from the SAAE well in Ehleringer, J.R., Bowen, G.J., Chesson, L.A., West, A.G., Podlesak, D.W., Cerling, T.E.,
2008a. Hydrogen and oxygen isotope ratios in human hair are related to geography.
the Distrito da Mina neighborhood and emerging again in the Proc. Natl. Acad. Sci. U.S.A. 105, 2788–2793.
Angolinhas upwelling zone. Ehleringer, J.R., Cerling, T.E., West, J.B., Podlesak, D.W., Chesson, L.A., Bowen, G.J.,
Integrated application of the methods allowed not only to elucidate 2008b. Spatial considerations of stable isotope analyses in environmental forensics.
In: Hester, R.E., Harrison, R.M. (Eds.), Issues in Environmental Science and
hydraulic communications but also all the dynamics of the meteoric, Technology. Royal Society of Chemistry Publishing, Cambridge, pp. 36–53.
surface and ground waters of the karst area. It is suggested the creation Feitosa, F.A., Manoel Filho, J., Feitosa, E.C., Demetério, J.G.A., 2008. Hidrologia: con-
of a permanent hydrochemical monitoring network through the wa- ceitos e aplicações, 3 ed. CPRM; LABHID, Rio de Janeiro, pp. 812 rev. e ampl.
Feng, W., Banner, J.L., Guilfoyle, A., Musgrove, M., James, E.W., 2012. Oxygen isotopic
tershed limits to identify the origin of contaminants such as nitrate,
fractionation between drip water and speleothem calcite: a 10-year monitoring study.
arsenic, cadmium, aluminum, manganese among others already iden- central Texas. USA. Chem. Geol. 304–305, 53–67.
tified in the area, as well as its dynamics in flow networks and its travel Filippini, M., Squarzoni, G., De Waele, J., Fiorucci, A., Vigna, B., Grillo, B., Stumpp, C.,
time; more detailed geophysical studies aiming to a better knowledge of 2018. Differentiated spring behavior under changing hydrological conditions in an
alpine karst aquifer. J. Hydrol. 556, 572–584.
the underground voids reported here, as well as their depths in relation Ford, D.C., Williams, P.W., 2007. Karst Geomorphology and Hidrology. Wiley, UK, pp.
to the surface and respective dimensions are suggested; systematic 562.
studies with dye tracers to confirm other hydraulic communications, as Foster, S., Hirata, R., Gomes, D., D'Elia, M., Paris, M., 2002. Groundwater Quality
Protection: a Guide for Water Service Companies, Municipal Authorities and
well as the construction of a Pains local meteoric isotopic line, will help Environment Agencies. World Bank, Washington DC, pp. 105. https://doi.org/10.
to improve interpretation of the overall karst region. 1596/0-8213-4951-1.
Freitas, P.M.C., 2009. Estudo de intercomunicação hidráulica superficial – subterrânea no
Carste da área periurbana de Pains/Mg, utilizando-se traçador Rodamina WT.
Declaration of competing interest Dissertação (mestrado). Universidade Federal de Minas Gerais, Instituto de
Geociências.
Galvão, P., Hirata, R., Halihan, T., Terada, R., 2017. Recharge sources and hydrochemical
This manuscript has not been published or presented elsewhere in evolution of an urban karst aquifer, Sete Lagoas, MG, Brazil. Environ. Earth Sci. 76
part or in entirety and is not under consideration by another journal. (4), 159.

12
T.N. Lucon, et al. Journal of South American Earth Sciences 100 (2020) 102591

Galvão, P., Hirata, R., Conicelli, B., 2018. Estimating groundwater recharge using GIS- aquifer vulnerability to contamination. J. S. Am. Earth Sci. 90, 191–203.
based distributed water balance model in an environmental protection area in the Penãranda, J.R., 2016. Condicionamento estrutural e litológico da porosidade cárstica da
city of Sete Lagoas (MG), Brazil. Environ. Earth Sci. 77 (10), 398. Formação Sete Lagoas, Município de Sete Lagoas (MG). Doctoral dissertation,
Geosoft, 2014. Calculating the Analytic Signal and Generalised and Tilt Derivates in Universidade de São Paulo.
MAGMAP. Pires, L.O., Rudnitzki, I.D., 2019. Análise de fácies e estratigrafia do maciço carbonático
Gibson, J.J., Edwards, T.W.D., Birks, S.J., St Amour, N.A., Buhay, W.M., McEachern, P., da Gruta do Éden, Pains - MG. In: Zampaulo, R.A. (Ed.), Congresso Brasileiro de
Wolfe, B.B., Peters, D.L., 2005. Progress in isotope tracer hydrology in Canada. Espeleologia, 35, 2019. Bonito. Anais. Campinas: SBE, 2019, pp. 63–72. Available in:
Hydrol. Process. 19 (1), 303–327. http://www.cavernas.org.br/anais35cbe/35cbe_063-072.pdf Accessed in: 30/01/
Goldscheider, N., Drew, D., 2007. Methods in Karst Hydrology. Taylor & Francis, London. 2019.
Greenberg, A.E., Clesceri, L.S., Eaton, A.D., 1995. In: Standard Methods for the Pizarro, A.P., 1998. Compartimentação Geológica Geomorfológica da Província
Examination of Water and Wastewater, vol. 18 American Public Health Association, Carbonática e Espeleológica de Arcos-Pains-Doresópolis. XL Congresso Brasileiro de
Washington, D.C. Geologia - BH – SBG.
Grohmann, C.H., Campanha, G.A.C., Soares Junior, A.V., 2011. OpenStereo: um pro- Plagnes, V., 1997. Le transport de matière dans les aquifères karstiques. In: 6th Conf on
grama Livre e multiplataforma para análise de dados estruturais. In: XIII Simpósio Limestone Hydrology and Fissured Media. La Chaux de Fond, Suisse 14-18/08/
Nacional de Estudos Tectônicos. 97:179-183.
Haddad, E.A., 2007. Influência antrópica na qualidade da água da bacia hidrográfica do Reis, H.L., Alkmim, F.F., Fonseca, R.C., Nascimento, T.C., Suss, J.F., Prevatti, L.D., 2017.
rio São Miguel, carste do alto São Francisco, Minas Gerais. Dissertação (Mestrado em The São Francisco basin. In: São Francisco Craton, Eastern Brazil. Springer, Cham, pp.
Geografia). Instituto de Geociências, Universidade Federal de Minas Gerais, Belo 117–143.
Horizonte, pp. 156. Ribeiro, J.H., Tuller, M.P., Danderfer Filho, A., 2003. Geological mapping of the region of
Haddad, E.A., Magalhães Jr., A.P., 2010. Influência antrópica na qualidade da água da Sete Lagoas, Pedro Leopoldo, Matozinhos, Lagoa Santa, Vespasiano, Capim Branco,
bacia hidrográfica do rio São Miguel, carste do alto São Francisco, Minas Gerais. Prudente de Morais, Confins and Funilândia, Minas Gerais State, Brazil (scale
V.25, n49, p79. Geosul 25 (49), 79–102. https://doi.org/10.5007/2177-5230.2010. 1:50,000), second ed. Belo Horizonte, pp. 54.
Han, Z., Tang, C., Wu, P., Zhang, R., Zhang, C., 2014. Using stable isotopes and major ions Ribeiro, A., Paciullo, F.V.P., Senra, A.S., Valeriano, C.M., Trouw, R.A.J., 2008. Geologia
to identify hydrological processes and geochemical characteristics in a typical karstic da Folha Piumhi- SF.23-V-B-II, escala 1:100.000: nota explicativa. UFRJ/CPRM,
basin. Guizhou. southwest China. Isot. Environ. Health Stud. 50 (1), 62–73. Minas Gerais 50p.
Harvey, F.E., Sibray, S.S., 2001. Delineating ground water recharge from leaking irriga- Roden, J.S., Lin, G.G., Ehleringer, J.R., 2000. A mechanistic model for interpretation of
tion canals using water chemistry and isotopes. Ground Water 39, 408–421. hydrogen and oxygen isotope ratios in tree-ring cellulose. Geochem. Cosmochim.
Heilbron, M., Cordani, U.G., Alkmim, F.F., 2017. São Francisco Craton, Earstern Brazil: Acta 64, 21–35.
Tectonic Genealogy of a Miniature Continent. Springer, pp. 326. Rozansky, K., Araguás-Araguás, L., Gonfiantini, R., 1993. Isotopic patters in modern
IAEA/WHO International Atomic Energy Agency/World Health Organization, 2004. global precipitation. In: Swart, P.K. (Ed.), Climate Change in Continental Isotopic
Global Network of Isotopes in Precipitation. The GNIP Database. Available at: Records. Geophysical Monograph Series 78. AGU, Washington, pp. 1–36.
http://isohis.iaea.org. Ruggieri, R., Forti, P., Antoci, M.L., De Waele, J., 2017. Accidental contamination during
Kaufmann, G., 2014. Geophysical mapping of solution and collapse sinkholes. J. Appl. hydrocarbon exploitation and the rapid transfer of heavy-mineral fines through an
Geophys. 111, 271–288. overlying highly karstified aquifer (Paradiso Spring, SE Sicily). J. Hydrol. 546,
Kaufmann, G., Romanov, D., Nielbock, R., 2011. Cave detection using multiple geophy- 123–132.
sical methods: unicorn cave, Harz Mountains, Germany. Geophysics 76 (3), B71–B77. Ryan, M., Meiman, J., 1996. An examination of short-term variation in water quality at
Killingley, J.S., Newman, W.A., 1982. 18O fractionation in barnacle calcite: a barnacle the karst spring in Kentucky. Ground Water 34 (1), 23–30. https://doi.org/10.1111/
paleotemperature equation. J. Mar. Res. 40, 893–902. j.1745-6584.1996.tb01861.x.
Klimchouk, A., Ford, D., 2000. Types of Karst and Evolution of Hydrogeologic Setting. Saadi, A., 1991. Ensaio sobre a morfotectônica de Minas Gerais: tensões intraplaca,
Speleogenesis. Evolution of Karst Aquifers. National Speleological Society, descontinuidades crustais e morfogênese. Thesis (Full Professor). Institute of
Huntsville, pp. 47–53. Geosciences, Federal University of Minas Gerais. Belo Horizonte, pp. 300.
Kresic, N., 2013. Water in Karst. Management, Vulnerability and Restoration. McGraw- Schoeller, H., 1962. Les Eaux Souterraines. Masson & Cie, Paris, pp. 642p.
Hill, New York, pp. 708. SEE - Sociedade Excursionista e Espeleológica, 2012. Projeto Arcos Pains Espeleologia –
Li, S.L., Liu, C.Q., Li, J., Lang, Y.C., Ding, H., Li, L., 2010. Geochemistry of dissolved PROAPE. Termo de Ajuste de Conduta (TAC) MPF/IBAMA/GERDAU/FEOP/SEE.
inorganic carbon and carbonate weathering in a small typical karstic catchment of Ouro Preto. Accessed February 2018: http://www.see.ufop.br/wp-content/uploads/
Southwest China: isotopic and chemical constraints. Chem. Geol. 277, 301–309. 2013/08/projeto-arcos-pains613 espeleologia.pdf.
Liu, L., Shu, L., Chen, X., Oromo, T., 2009. The hydrologic function and behavior of the Smith, B., Sandwell, D., 2003. Accuracy and resolution of shuttle radar topography
Houzhai underground river basin. Guizhou Province. southwestern China. Hydrogr. mission data. Geophys. Res. Lett. 30 (9), 1467–1470.
J. 18, 509–518. Somaratne, N., Mann, S., 2016. Integrated use of geological, geophysical, radiocarbon and
Lucon, T.N., Costa, A.T., Galvão, P., Leite, M.G.P., 2018. Natural background levels and stable isotopes data for tracing the conduit flow paths in a small karstic aquifer:
seasonal influence on groundwater chemistry of the Upper São Francisco karst region, poocher swamp freshwater lens, south Australia. Environ. Nat. Resour. Res. 6 (3),
MG, Brazil. Braz. J. Genet. 48 (4), 867–879. 119.
Luo, M., Chen, Z., Zhou, H., Jakada, H., Zhang, L., Han, Z., Shi, T., 2016. Identifying Souza, E., Galvao, P., Almeida, R., Pinheiro, C., Baessa, M., Cabral, M., 2015. Stable
structure and function of karst aquifer system using multiple field methods in karst isotopes studies in the urucu oil Province, amazon region, Brazil. J. Water Resour.
trough valley area, South China. Environ. Earth Sci. 75 (9), 824. Protect. 7, 131 03.
Martins, T.I.S., 2013. Mapeamento Geomorfológico da Folha Piumhi, Minas Gerais. 157f. Teixeira, C.M., Duque, T.R., Rosa, M.L., Renó, R., Lucon, T.N., Aquino, I.B., Santos, T.F.,
Dissertação (Mestrado) Universidade Federal de Uberlândia Disponível em: http:// 2013. Domínios espeleogenéticos da região de Arcos-Pains-Doresópolis MG, Brasil.
www.bdtd.ufu.br/tde_busca/arquivo.php?codArquivo=4856. In: Congresso Brasileiro de Espeleologia, vol. 32. pp. 381–390.
Menegasse, L.N., Gonçalves, J.M., Fantinel, L.M., 2002. Disponibilidades hídricas na Valeriano, M.D.M., Rossetti, D.D.F., 2008. TOPODATA: Seleção de coeficientes
província cárstica de Arcos-Pains-Doresópolis, alto São Francisco, Minas Gerais, geoestatísticos para o refinamento unificado de dados SRTM. INPE, São José dos
brasil. Revista Águas Subterrâneas 16 (1), 1–19. Campos.
Minardi, P.S.P., Bomtempo, V.L., 2000. Traçadores e Técnicas isotópicas em Hidrologia Weight, W.D., Sonderegger, J.L., 2000. Manual of Applied Field Hydrogeology. McGraw-
Subterrânea: A Experiência do CDTN/CNEN. Águas Subterrâneas, Brasil. Hill, New York, pp. 608.
Mobus, G., 2003. Qualigraf: software para interpretação de análises físico-químicas, White, W.B., 2003. Conceptual models for karstic aquifers. Speleogenesis and Evolution
versão Beta. Fundação Cearense de Meteorologia e Recursos Hídricos – FUNCEME, of Karst Aquifers 1 (1), 2.
Fortaleza. Williams, P., 2004. Encyclopedia of Caves and Karst Science, second ed. Fitzroy Dearborn
Mourão, M.A.A., Cruz, W.B., Gonçalves, R.L.F., 2001. Caracterização Hidrogeológica da Publ., Ch. Dolines, pp. 304–310.
Porção Mineira da Bacia Hidrográfica do São Francisco. In: Pinto, C.P., Martins-Neto, Wong, C., Kromann, J.S., Hunt, B.B., Smith, B.A., Banner, J.L., 2013. Investigating
M.A. (Eds.), Bacia do São Francisco Geologia e Recursos Naturais, Belo Horizonte, groundwater flow between edwards and trinity aquifers in central Texas.
SBG/MG, pp. 327–350. Groundwater 52, 624–639 10. 1111/gwat.12106.
Mudry, J., 1987. Apport du traçage physico-chimique naturel à la connaissance hydro- Worthington, S.H.R., 2005. Hydraulic and geological factors influencing conduit flow
cinematique des aquifères carbonates. Mémoire n4 Univ. Besançon. depths. Cave Karst Sci. 31 (3), 123–134.
Muzzi-Magalhães, P., 1989. Analise Estrutural Qualitativa das Rochas do Grupo Bambuí Yepez, E.A., Williams, D.G., Scott, R.L., Lin, G., 2003. Partitioning overstory and un-
na porção Sudoeste da Bacia do São Francisco. MSc Thesis, Univ. Ouro Preto, Brasil. derstory evapotranspiration in a semiarid savanna wood land from the isotopic
Palmer, A.N., 1991. Origin and morphology of limestone caves. Geol. Soc. Am. Bull. composition of water vapor. Agric. For. Meteorol. 119, 53–68.
103, 1–2. Zhao, M., Zeng, C., Liu, Z., Wang, S., 2010. Effect of different land use/land cover on karst
Pereira, D.L., Galvão, P., Lucon, T.N., Fujaco, M., 2019. Adapting the EPIK method to hydrogeochemistry: a paired catchment study of Chenqi and Dengzhanhe. Puding.
Brazilian Hydro(geo)logical context of the São Miguel watershed to assess karstic Guizhou. SW China. J. Hydrol 388, 121–130.

13