View of Hypogean microclimatology and hydrology of the 800-900 m asl level in the Monte Corchia cave (Tuscany, Italy): preliminary considerations and implications for paleoclimatological studies

(1)

HYPOGEAN MICROCLIMATOLOGY AND HYDROLOGY OF THE 800-900 m ASL LEVEL IN THE MONTE CORCHIA CAVE (TUSCANY, ITALY): PRELIMINARY CONSIDERATIONS AND IMPLICATIONS FOR IMPLICATIONS FOR

PALEOCLIMATOLOGICAL STUDIES

MIKROKLIMA IN HIDROLOGIJA NA NADMORSKI VI�INI 800-900 m V JAMSKEM SISTEMU MONTE CORCHIA: PRVE UGOTOVITVE IN

POMEN ZA PALEOKLIMATOLO�KE RAZISKAVE

Ilaria BANESCHI¹, Leonardo PICCINI²,Eleonora REGATTIERI², Ilaria ISOLA⁴, Massimo GUIDI¹, Licia LOTTI⁵, Francesco MANTELLI⁶, Marco MENICHETTI⁷, Russell N. DRYSDALE⁸ & Giovanni ZANCHETTA^1,3,4

Izvleček UDK 556.3+551.584(450.52)

Ilaria Baneschi, Leonardo Piccini, Eleonora Regattieri, Ilaria Isola, Massimo Guidi1, Licia Lotti, Francesco Mantelli, Marco Menichetti, Russell N. Drysdale & Giovanni Zanchetta: Mikro- klima in hidrologija na nadmorski višini 800-900 m v jam- skem sistemu Monte Corchia: prve ugotovitve in pomen za paleoklimatološke raziskave

Jamski sistem Monte Corc�ia je ena najobetavnejši� krajev za paleoklimatske raziskave v Sredozemlju. Kljub temu je �idrologija in �idrokemija sistema slabo poznana. V članku pred- stavimo nekatere klimatsko-meteorološke in �idrokemične podatke iz različni� delov sistema. Meritve prevodnosti in vodnega nivoja v kanalu La Gronda, kažejo �iter odziv sistema na padavine in dobro razvito epikraško cono. Odziv prevodnosti, pH, in δ¹³CDIC v curki� kaže na lokalno strukturo in dolžino vodne pot vode skozi vadozno cono. V Geleriji stalaktitov, ki je osrednje območje paleoklimatski� raziskav, smo zabeležili stabilno temperaturo, specifično električno prevod- nost, pH, alkalnost ter koncentracijo kalcija, magnezija in δ¹⁸O.

Hitrost kapljanja se tu ne odziva na padavine, ampak kaže dol- go časovne trende, kateri� značilnosti bomo la�ko videli šele po daljšem opazovanju. Trenutni podatki podpirajo �ipotezo, da je Galerija Stalaktitov dober primer globokega podzemnega sistema v smislu, kot ga definira Fairc�ild (2007).

Ključne besede: kraška �idrologija, jamska klima, podzemna voda, kapniki, Antro del Corc�ia, Apuanske Alpe.

1 Istituto di Geoscienze e Georisorse – CNR, Via Moruzzi, 1, 56100, Pisa, Italia, e-mail: i.banesc�i@igg.cnr.it

2 Dipartimento di Scienze della Terra, Università di Firenze, Via La Pira, 4, 50121, Firenze, Italia

3 Dipartimento di Scienze della Terra, Università di Pisa, Via S. Maria 53, 56126 Pisa, Italia

4 Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Pisa, Via della Faggiola 32, 56126 Pisa, Italia

5 Agenzia Regionale per la protezione Ambientale della Toscana (ARPAT) – Dipartimento Provinciale di Massa, Via del Patriota, 2, 54100 Massa, Italia.

6 Agenzia Regionale per la protezione Ambientale della Toscana (ARPAT) – Dipartimento Provinciale di Firenze, Via Ponte alle Mosse, 211, 50144 Firenze, Italia.

7 Dipartimento di Geologia e Tecnologie C�imic�e Ambientali Università di Urbino Campus Universitario I-61029 Urbino, Italy

8 Sc�ool of Environmental and Life Sciences, University of Newcastle, Callag�an, NSw 2308, Australia Received/Prejeto: 15.9.2010

Abstract UDC 556.3+551.584(450.52)

Ilaria Baneschi, Leonardo Piccini, Eleonora Regattieri, Ilaria Isola, Massimo Guidi, Licia Lotti, Francesco Mantelli, Marco Menichetti, Russell N. Drysdale & Giovanni Zanchetta: Hypo- gean microclimatology and hydrogology of the 800-900 m asl level in the Monte Corchia cave (Tuscany, Italy): Preliminary considerations and implications for paleoclimatological studies The Monte Corc�ia Cave is one of t�e most promising sites for studying t�e paleoclimate of t�e Mediterranean basin, but its �ydrology and �ydrogeoc�emistry are still poorly known.

In t�is paper, we report some meteoclimatic and �ydroc�emical data for different parts of t�e cave. Conductivity and water level data from La Gronda c�annel s�ow t�at t�is system reacts rapidly to external meteoric events, indicating t�e presence of a conductive epikarst. Data on two different drips indicate t�at t�e p�ysicoc�emical parameters, suc� as conductivity, pH, δ¹³CDIC and drip rate depend on t�e local structural setting and water pat� lengt�. The data presented s�ow t�at Galleria delle Stalattiti (t�e focus of t�e paleoclimate researc�) �as t�e most stable conditions in terms of temperature, and t�e dripwaters s�ow constant pH, electrical conductivity, alkalinity, calcium and magnesium content and δ¹⁸O. Drip rate is not affected by rain events and displays long-term trends t�at require a longer period of monitoring for elucidating t�eir nature. The preliminary data presented �ere corroborate t�e �ypot�eses suggesting Galleria delle Stalattiti as a good example of a “deep” �ypogean system of Fairc�ild et al. (2007).

Keywords: karst �ydrology, cave meteoclimatology, groundwa- ter, speleot�ems, Antro del Corc�ia, Alpi Apuane.

(2)

In recent years, t�e Monte Corc�ia Cave system (Alpi Apuane, central Italy) �as turned out to be one of t�e most promising arc�ives for reconstructing t�e paleoclimate and t�e paleoenvironment in t�e Mediterranean basin (Drysdale et al. 2004, 2005, 2007, 2009; Zanc�etta et al. 2007), wit� implications for t�e c�ronology of climatic events at t�e global scale (Drysdale et al. 2007, 2009). However, t�e �ydrology of t�is large, �ypogean system is still far from being completely understood, and t�e information from climate proxy data extracted from Corc�ia speleot�ems is far from complete.

Recently, Fairc�ild et al. (2007) suggested a classifi- cation of �ypogean environments based on t�e relation- s�ip between flow pat�, drip c�emistry and cave climate, and cave climate, w�ic� produces different textures and compositions of speleot�ems. Specifically, t�ey describe a deep environment c�aracterized by stable conditions in terms of temperature and �umidity wit� slow growt� of speleot�ems, free of influence by s�ort-term (seasonal) variations.

They suggested t�at suc� conditions could be present only in very deep caves far from t�e surface, but t�ey do not report examples.

Piccini et al. (2008) �ave recently suggested t�at t�e Galleria delle Stalattiti in Corc�ia Cave mig�t represent an example of t�e uniform �ypogean case of Fairc�ild et al. (2007). However, t�e data publis�ed were rela- tively sparse and influenced by t�e fact t�at t�e size and remoteness of t�is cave �as prevented t�e acquisition of

accurate and sufficiently long monitoring data from t�e part of greatest scientific interest before t�e opening of t�e cave to tourism. Considering t�e exceptional c�aracteristics of t�is site and t�e relevance of paleoclimatic implications of t�e researc�, more detailed monitoring

�as started recently to better c�aracterize t�is cave environment. This paper discusses some of t�e preliminary results from t�is work. However, differently from Pic- cini et al. (2008), w�o described t�e general features of t�e cave system, in t�is paper we focus on a preliminary comparison of t�e basic �ydro-geoc�emical and �ydrological c�aracteristics of different sub-systems wit�in Corc�ia: La Gronda stream and two drips, located at about t�e same elevation range (from 800 to 900 m asl), but in two different sectors of t�e cave c�aracterized by different geological and structural (e.g. rock t�ickness above t�e cave) contexts. This study points out t�e com- plexity of t�is cave and �ig�lig�ts t�e particular climatic stability of t�e Galleria delle Stalattiti. The data presented

�ere support t�e general idea t�at Galleria delle Stalat- titi could be a good example of a “�ypogean cave” sensu Fairc�ild et al. (2007).

As a deep environment wit� low drip rate wit� continuous flow and constant c�emistry in a long time scale, t�e Galleria delle Stalattiti offers t�e possibility of a more detailed understanding of t�e landscape, tectonic and climatic evolution of t�e region, as introduced by Dry- sdale et al. (2004).

INTRODUCTION

STUDY AREA

The Monte Corc�ia karst system is located in t�e sout�- western part of t�e Alpi Apuane (Nw Tuscany, Italy) (Fig. 1) and is developed inside t�e carbonate core of a syncline, almost completely enclosed by a non-karstifi- able, low permeability basement. Corc�ia Cave is about sixty km long and 1197 m deep and it is presently t�e largest and one of t�e deepest caves in Italy. The presently known cave consists of a 3D network of passages distrib- uted from 1550 to 450 m asl, �aving t�e s�ape of regular tubes or canyons and wit� a mainly �orizontal pattern.

Several �ig�-gradient passages intersect epip�reatic net- works along t�e main fractures and form t�e present active drainage system from t�e surface to t�e main collec- tor (Piccini 1998). The conduit level located between 900 and 800 m asl is still active in t�e inner part of t�e system, w�ereas it forms a relict level of p�reatic passages t�at

�ave been part of t�e tourist pat� since 2001. The system now �as 14 entrances accessible to �umans.

The Alpi Apuane represents a tectonic window t�at s�ows t�e deepest exposed levels (Tuscan Metamorp�ic Units) of t�e Nort�ern Apennines (Carmignani & Klig- field 1990; Molli & Vaselli 2006 and references t�erein).

The Mesozoic cover includes Triassic continental to s�allow water deposits followed by Upper Triassic-Li- assic carbonate platform metasediments (“Grezzoni”

dolostones, dolomitic marbles and marbles), w�ic� are succeeded by Upper Liassic-Lower Cretaceous c�erty metalimestones, c�erts and calcsc�ists, and Lower Cre- taceous to Lower Oligocene sericitic p�yllites and calcsc�ists. The Oligocene-Early Miocene sedimentation of turbiditic metasandstone closes t�e sedimentary �istory of t�e domain.

(3)

METHODS

MICROCLIMATOLOGY

At Corc�ia Cave, microclimate monitoring by ARPAT (Agenzia Regionale per la Protezione dell’Ambiente, To- scana) started in 1998. Three microclimatic stations were installed along t�e tourist pat� (Fig. 2). Station ARPAT-1 is at t�e intersection of t�e touristic artificial tunnel and t�e natural pat� (870 m asl); station ARPAT-2 is located at Galleria del Venerdì, about 1.4 km from t�e entrance at 830 m asl; station ARPAT-3 is located at Galleria delle Stalattiti, at 840 m a.s.l. and about 2 km from t�e artificial entrance, but data for t�is station are available only from 2002 to 2004. Due to tec�nical problems, only temperature data are validated, and even so, are discontinuously.

The sensor for temperature is a Pt-100 type wit� a measurement range of -50 to +80°C and a digital resolution of 0.1°C. Data are collected �ourly and stored in a datalogger. The University of Urbino placed a temperature sensor in t�e Galleria delle Stalattiti for t�e period 2005- 2008 (�ere, named URB-2, Fig. 2).

Two microclimate stations, measuring temperature, precipitation, wind speed and atmosp�eric pressure, �ave been installed on t�e slopes of Mt. Corc�ia at two different elevations. The stations include a tipping-

bucket rain gauge wit� a resolution of ±0.2 mm/� and accuracy of 0.1 mm/� for events of 10 mm/� and 0.3 mm/� for events of 60 mm/�, an air temperature sensor wit� a range of measurement of -50 to +80°C and a°C and aC and a digital resolution of 0.1°C, an atmosp�eric pressure sen-°C, an atmosp�eric pressure sen-C, an atmosp�eric pressure sensor wit� a resolution and accuracy of ±0.5 mbar and a range of measurement of 600 to 1100 mbar, and a wind speed sensor wit� an accuracy of 0.1 m/s between 0 and 1 m/s and 0.25 m/s between 1 to 10 m/s and a range of measure of 0 to 50 m/s. All stations take �ourly readings of t�e variables and store t�ese data in a datalogger. The first station (MT-1) is at 860 m asl, just at t�e entrance of t�e touristic artificial tunnel, t�e second one (MT-2) at 1480 m asl, near t�e �ig�est western crest of t�e mountain. Because of t�e position of MT-2, t�ere is a loss of rain w�en t�e wind blows from west wit� ascending mo- tion. After comparison wit� ot�er rain stations around t�e studied area, rainfall measurements at MT-2 were not considered reliable, and t�ey are not included in t�is work. Unfortunately, records of MT-1 are not continuous because of tec�nical problems due to t�e severe microclimatic conditions. For t�is reason, �ere we present only significant and validated time series of data. In t�is work, The alpine-like landscape of t�is peculiar region is

due to t�e complex structural setting, c�aracterized by Fig. 1: Location map of monte Corchia Cave.

very steep bedding and by t�e tectonic repetition of different rocks (mainly limestone and p�yllite), w�ic� �as en-

�anced t�e role of differential erosion. The present topogra- p�y is t�e result of �eavy flu- vial erosion accompanied by rapid tectonic uplift during t�e Early Pleistocene (Piccini et al. 2003). Glacial and peri- glacial processes res�aped t�e landscape, emp�asizing t�e “alpine-like” features of t�is mountain range (Bras- c�i et al. 1986). On limestone rock, only a few large-scale karst landforms occur, w�ile medium and small-scale landforms are widely repre- sented. In any case, t�e most important karst features of t�is area consist of several cave systems among w�ic�

Corc�ia is t�e largest.

(4)

we illustrate data monitored in different parts of Corc�ia system during t�e period 2005-2009. Tab. 1 summarizes t�e location and description of t�e stations.

Air temperature measurements at t�e CNR-1 station started in September 2009; t�erefore, t�ese data are preliminary. The data logger (Escort Data Loggers Ltd) records temperature (accuracy 0.3°C) every 10 minutes, and t�e data are downloaded using ESCORT Console software.

HYDROLOGY

Since December 2008, a �ydrological station (named IDRO-1) �as been installed along t�e tourist pat�way at La Gronda C�annel, t�e main active streamway close

to t�e Galleria delle Stalat- titi (Fig. 2). The instrument consists of a multi-parameter probe provided wit�

a t�ermometer, a pressure sensor for water level measurements, a conductivity-meter, a dissolved oxygen sensor, a pH-meter and a turbidity sensor. The sensors are connected to a data logger t�at collects and processes t�e measurements. The monitoring covers only five mont�s up to April 2009, due to t�e damaging of t�e instruments during a strong flood. Nevert�eless, t�e record furnis�es a first set of data concerning t�e �ydrological be�avior of a stream near t�e Galleria delle Stalattiti c�amber during t�e winter season.

This sensor set was placed in a small artificial pool, so water level is affected by t�e t�res�old effect due to t�e artificial damming of t�e water flow. Therefore, water level record needs to be viewed wit� caution during �ig�

disc�arge events because of turbulence.

Fig. 2: Detailed map of the monte Corchia, 800-900 m cave level, with drip and microclimatic sta- tions.

(5)

Drip rate counters �ave been placed under a sta- lactite since October 2007 at station CNR-2 and since May 2009 at station CNR-1 (Fig. 2). Drip counters used in t�is study are t�e STALAGMATE recorders (Collis-

ter & Mattey 2008) (furt�er information can be found at

�ttp://81.174.171.35/Driptyc�/Stal_folder/Stal_frame- set.�tml) t�at acoustically generate electrical impulses of drops falling onto t�e lid of a watertig�t box, w�ic�

contains a signal detector board and a Gemini event data logger (�ttp://www.geminidataloggers.com). Eac� drip rate up to t�e maximum (5 drips/sec) is recorded, wit�

a storage interval of 30 minutes. Eac� drip counter was placed on t�e top of t�e dripwater sampling device; t�e dripwaters are collected into a tank by a funnel after contact wit� t�e drip counter. Bot� devices are still recording in t�e cave.

HYDROPHYSICAL ANALYSIS

In t�e last twelve years, several p�ysical and c�emical analyses of rain, surface and cave waters �ave been discontinuously collected by ARPAT and t�e CNR (Pisa) (Mantelli et al. 2005; Piccini et al. 2008). Most of t�e cave waters �ave been collected more recently along t�e tourist pat� and t�e most visited parts of t�e cave, alt�oug�

some samples �ave been collected in t�e upper sections of t�e karst system.

Mont�ly monitoring of dripwater c�emistry started in April 2009. The results presented �ere referred to data collected from April to September 2009 from t�e drips CNR-1 and CNR-2. Dripwaters are collected in a single bottle of eig�t litres for about one mont�, and t�e p�ysical and c�emical parameters analyzed on a composite sample.

water temperature, conductivity and pH are measured in t�e field by a portable multi-parameter data logger (DeltaOHM Instruments). Accuracy is 0.5% for con-

ductivity, 0.25% for temperature and 0.1 for pH. All of t�e sensors are calibrated on eac� sampling trip. Total alkalinity was measured in t�e field by HCl titration ac- cording to Gran’s met�od (Gran 1952) using met�yl-or-

ange as indicator.

Upon collection, eac�

sample was split into two aliquots. One aliquot was fil- tered and acidified wit� HCl for cation analysis (Ca, Mg, K and Na) in order to prevent calcite precipitation. The ot�er un-acidified aliquot was analysed for anions and stable isotopes. All solutions were stored in c�emically inert plastic bottles and kept in a cool bag until t�e end of t�e monitoring period prior to return to t�e laboratory for analysis. Cations were analysed by Atomic Absorption Spectrometry and anions by ion c�romatograp�y (DIONEx-100). The precisionion c�romatograp�y (DIONEx-100). The precisionThe precision for cations and anions were ≤±2%. Calcite saturation index (defined as t�e log of t�e quotient of ionic activity product and solubility product) and t�e partial pressure of CO₂ in equilibrium wit� t�e solution – expressed as log(pCO₂) – were determined from measured dripwater c�emistry on balanced analyses using t�e modelling program SOLVEq (Reed 1982).

The total inorganic carbon (TIC) of t�e waters was calculated from Total Alkalinity (TAlk), pH, ionic strengt� and temperature using t�e stoic�iometric ap- proac� of inorganic carbon equilibrium (Stumm &

Morgan 1981). Because of t�e low Total Organic Car- bon (TOC) concentrations (analyses of different waters from t�e cave system were between 0.1 mg C/l and 2 mg C/l, Mantelli et al. 2005; Piccini et al. 2008), t�e influence of organic anions on alkalinity was considered negligible.

The δ¹⁸O of t�e water was determined after equilibration wit� CO₂ (Epstein & Mayeda 1953). The isotopic measurements were made on a ThermoFinnigan MAT-252 mass spectrometer and are given in per mil values (‰) wit� respect to t�e V-SMOw standard (Craig 1961). Analytical reproducibility of duplicates measured on different days is better t�an 0.1‰. The δ¹³C of t�e total dissolved CO₂ in t�e water (DIC) was measured by adding dry H₃PO₄ to t�e water sample. The released CO₂ was t�en collected in a vacuum line, and t�e δ¹³C values were measured wit� a mass spectrometer as for oxygen and are given in ‰ wit� respect to t�e PDB standard (Craig 1957).

Tab. 1: Location and characteristics of monitoring sites.

Station

name Type Location Elevation

m asl

MT-1 Microclimatic Tourist cave entrance 850

MT-2 Microclimatic (only temperature

is considered in this work) SW flank of Mt. Corchia 1600 ARPAT-1 Microclimatic Cave at artificial tunnel end 875 ARPAT-2 Microclimatic Cave at 1.4 km from entrance 840 CNR-1 Drip and water sampling Cave at 0.80 km from entrance 860 CNR-2 Drip and water sampling Cave at 2 km from entrance 870

URB-2 Temperature Cave at 2 km from entrance 870

IDRO-1 Multiparameter hydrological Cave at La Gronda channel 830

(6)

AIR TEMPERATURE

Fig. 3 s�ows t�e temperature fluctuations of t�e two external stations (MT-1 and MT-2) and t�ree inner stations (ARPAT-1, 2 and URB-2) during t�e year 2005. Typical fluctuations emp�asize t�e extreme stability of stations ARPAT-2 and URB-2 compared to ARPAT-1, as expect- ed for t�e deepest sectors of t�e cave. Station ARPAT-1

�as a mean annual temperature significantly lower t�an t�e ot�ers two, due to t�e fact t�at it is located near two lower entrances t�roug� w�ic� cold air enters during winter, w�ile ARPAT-2, w�ic� is more t�an 1 km from t�e entrances, is only marginally influenced by external temperature. In fact, t�e maximum fluctuation in temperature for station ARPAT-1 is 4°C, from a minimum of 3.6 to a maximum of 9.3°C. For ARPAT-2, t�e temperature ranges from 7.6 to 9°C.

The station URB-2 at Galleria delle Stalattiti is relatively stable in terms of temperature, in agreement wit�

its position nearly 2 km from t�e nearest natural entrances. Temperature ranges from a minimum of 7.3 °C (in Marc�) to a maximum of 8.1°C (in October and No- vember).

CHANNELLED wATER FLOw

During t�e period of available measurements, t�e water level fluctuates between a maximum value of 70.1 cm and a minimum of 52.1 cm; t�e values of conductivity (EC) range between a maximum of 242 µS/cm and a minimum of 104 µS/cm. The temperature of t�e water was quite stable wit� amplitude of 0.8°C in t�e range of 6.9°C to 7.7°C, and a mean value of 7.3°C. The data recorded by pH probe, as illustrated in Fig. 4, denote t�e presence of peaks, s�ort in time (about 5

�ours) w�ere t�e signal sud- denly increases by ca. 1 pH unit. As t�ese large pH variation events are not associated wit� variations in water level, conductivity or temperature, t�ey could be considered an artifact of a malfunctioning sensor. Hence, t�ey are not included in t�e following statistical elaboration and discussion. The maximum value of pH was 7.5 and t�e minimum 6.

water level is rapidly influenced by rain events, as illustrated in Fig. 4 w�ere t�e be�avior of t�e water flow regime was compared to rainfall. This be�avior reveals t�e occurrence of an epikarst wit� a very good permeability. The negative correlation between electrical conductivity and water level, during t�e main storm events, suggests t�at t�e cave stream is fed by infiltration t�roug�

vadose passages wit� rapid flow rate. The pH decreases in correspondence to t�e rainfall events, indicating water wit� a lower pH.

Fig. 3: Temperature at the station ARPAT-1, ARPAT-2, URB-2, mT-1 and mT-2 during winter- summer 2005. The temperature data of ARPAT-1, mT-1 and mT-2 are plotted on the left y-axis (L); while URB-2 and ARPAT-2 are on the right y-axis (R).

Fig. 4: Water level, electrical conductivity and ph of gronda stream (IDRO-1) compared with rainfall.

RESULTS

(7)

DRIP RATES

The disc�arge at t�e CNR-1 and CNR-2 stations did not cease during t�e studied period. The Fig. 5 under- lines t�e different be�avior of t�e two drips and t�eir

different sensitivities to rain events. CNR-1 is fed by a fracture c�aracterized by a rapid response to infiltration. Drip rate ranges from about 60 drips/� to more t�an 7000 drips/�. The s�ort period for monitoring does not yet allow us to resolve significant correlations and time periodicity. CNR-2 s�ows a different be�avior: drip counting ranges from 60 to 130 drips/�, except for a s�ort period (from 23 and 24 September 2009) w�en it reac�ed about 450 drips/�. This exceptional event lasted about 12 �. However, as t�e event �appened only once during t�e monitoring period, we cannot say if it is due to a genuine be�avior of t�e system or was caused by an electrical disturbance.

In t�e CNR-2 drip record, t�ere is not a simple and direct correlation wit� rainfall events as recorded at t�e entrance. It is clear, in fact, t�at t�ere is no rapid clear, in fact, t�at t�ere is no rapid response of drip rate to rain events, because t�e relative maxima of t�e drip rate in August and October are not coeval wit� rainy periods. Fig. 6 s�ows a c�ange ins�ows a c�ange in t�e drip rate during t�e monitoring period. In particular, for t�e year 2008, a descending trend is recorded up to 3 September 2008, w�en t�e drip rate increased in a few �ours from 80 to 90 counts/� and t�en started to Fig. 5: Drip rate at the stations CNR-1 and CNR-2 and hourly rainfall at the station ARPAT-1, near the tourist entrance.

reduce again. On 8 January 2009 t�e drip rate reac�ed 125 counts/� and descended slowly.

The period of t�e monitoring is s�ort to allow meaningful time series analyses. Nonet�eless, as a first

attempt, seasonal fluctuations are clearly not evident.

Furt�er, t�e first long descending trend up to 3 Septem- ber 2008 is reproduced by an exponential curve, w�ereas t�e second and t�irds events are better described by a parabolic curve. This suggests t�at t�e be�avior of t�e system during t�ese two events can be reproduced by an equation simulating a piston flow-disc�arging system. A longer monitoring period is needed to model t�e �ydrology of t�e drip.

HYDROPHYSICS AND HYDROCHEMISTRY water c�emistry of dripwaters at t�e CNR-1 and CNR-2 sites from April to September 2009 is illustrated in Fig. 7. The two stations s�ow differences in t�e �ydro- c�emical composition and in t�e variance of a number of parameters. Tab. 2 s�ows t�e mean and t�e standard deviation (SD) for t�e measured parameters. Hydroc�emical parameters t�at do not s�ow variability t�roug�out t�e monitoring period at t�e station CNR-2 include Ca, Mg, alkalinity and conductivity. The pH varies at least from a maximum of 8.41 to a minimum of 8.10, but ir- regularly, wit�out any significant correlation wit� drip rate. Also water temperature increases in a non-mo- notonous way from 7.6 to 8.0°C. The partial pressure of CO₂ is mainly invariant wit� a peak in July; in addition saturation index for calcite is constant wit� a mean value of 0.4 and a minimum in July. The Ca, Mg and alkalinity content is mainly invariant also at t�e station CNR-1.

The differences between CNR-1 and CNR-2 are evident, not only in t�e Ca and Mg content of t�e two dripwaters Fig. 6: Drip rate at the station CNR-2 in relation to rainfall. Data are hourly.

(8)

but also in Mg/Ca molar ratio t�at is 1.21 for CNR-2 and 0.36 for CNR-1.

CNR-1 is c�aracterized by a lower conductivity and alkalinity t�an CNR-2 (Tab. 2 and Fig. 8). water temperature at CNR-1 �as a �ig�er variability, wit� a relative minimum in April and June and an increase from June to September. Indeed, t�e interquartile range for t�e temperature is 0.3°C and 0.7°C for CNR-2 and CNR-1, respectively. Drip site CNR-1 s�ows a quite regular variation of t�e pH wit� a minimum during July. Conse- quently saturation index (SI) for calcite in t�is dripwater is lowest during late July. Hence, bot� pH and SI s�ow a trend t�at mirror t�e calculated pCO₂ of t�e waters.

we are aware t�at a one mont�-aggregate sample s�ould undergo compositional c�anges as a result of

Fig. 7: Water chemistry of CNR-1 and CNR-2 stations from April to September 2009.

Tab. 2: Statistics (mean and standard deviation, SD) for chemical composition of the dripwaters collected for one year (2009-2010) for a total of 10 samples.

CNR-2 CNR-1

Mean SD Mean SD

Water Temperature (°C) 7.8 0.2 7.9 0.4

Conductivity ( µS/cm, 25 °C) 342.6 8.5 228.6 7.9

pH 8.28 0.11 8.16 0.11

Alkalinity mmol/l 2.74 0.08 2.15 0.09

HCO₃ (mmol/l) 2.74 0.09 2.15 0.07

Ca mmol/l 0.78 0.01 0.80 0.02

Mg mmol/l 0.94 0.01 0.29 0.01

equilibration wit� t�e cave atmosp�ere and calcite precipitation. The CO₂ degassing of dripwater causes an increase of t�e pH wit� a consequent rise in carbonate species (as CO₃^--) dissolved in waters. In t�ese conditions calcite precipitation s�ould be t�ermodynamically fa- voured alt�oug� t�e kinetics of precipitation is slow (Drey- brodt et al. 1997). Moreo- ver, low Ca concentration in dripwater would severely dampen calcite precipitation rates (Genty et al. 2001). As

�ig�lig�ted by Drysdale et al. (2004), it was not possible to collect fres� calcite for bot� t�e CNR-1 and CNR-2 dripwater sites. In t�is condition a potential increase of pH due to CO₂ degassing and t�e oversaturation state s�ould take place in a condition of constant alkalinity and Ca. To investigate w�et�er a c�emical-p�ysical reaction �appens, we �ave started monitoring t�e pH of t�e drip “instantaneously”

and collecting aliquots of water for some �ours, in order to evaluate differences due to 1 mont� of storage.

Extensive literature (Baker et al. 1997, 2000;

Musgrove & Banner 2004; Spötl et al. 2005; Baldini et al.

2006; McDonald et al. 2007) indicates t�at, wit�in t�ewit�in t�e same cave c�amber, �ig�ly conductive flow and slow, constant flow can co-exist, and t�ese different drips can

�ave substantially different c�emical and p�ysical prop- erties. However, t�is is probably not t�e case at t�e Gal- leria delle Stalattiti, w�ic� is under ca. 400 m of rock and is located just below t�e impermeable basement (see also Piccini et al. 2008), making it difficult to imagine significant differences among drips. The selected site resulted from a preliminary selection of ot�er sites (even if t�ere are relatively few drips in t�e c�amber), s�owing a very similar be�aviour (obtained only by scattered observa- tion), and it was selected considering accessibility and to avoid disturbing t�e tourist pat�.

(9)

The oxygen isotope composition at t�e drip station CNR-2 was significantly constant during t�e monitoring period as illustrated in Fig. 9, confirming t�e data discussed by Piccini et al. (2008). The drip station CNR-1 s�ows relatively more variability wit� a lower value measured in April. Piccini et al. (2008) also measured t�e most negative value in April.

This could confirm t�at CNR-1 is likely influenced by seasonal effects and in t�is case by infiltration of snow melt c�aracterized by more negative isotopic values. The carbon isotope composition of t�e dissolved inorganic carbon (DIC) is quite different for t�e two stations, and CNR-1 (δδ¹³C = -6.66±0.52) = -6.66±0.52) s�ows more variability t�an CNR-2 (δδ¹³C = -3.36±0.15), = -3.36±0.15), wit� low values in April and July.

Fig. 9: Oxygen and carbon iso- tope composition of dripwaters at the stations CNR-1 and CNR-2.

Fig. 8: Box-whisker for temperature, ph, conductivity and alcalinity of the dripwater collected at the stations CNR-1 and CNR-2. For the elaboration 15 samples were processed for CNR-2 and 11 for CNR-1.

DISCUSSION

The �ydrological and �ydrop�ysical monitoring of t�e 800-900 m asl cave level in Mt. Corc�ia s�ows very different conditions for t�e two sectors of t�e cave considered.

The karst system �as, in general, a very conductive epikarst as indicated by t�e response of t�e La Gronda c�annel to storms. An inverse correlation between EC and disc�arge indicates t�at t�is streamway is fed by in-

filtration t�roug� vadose passages wit� rapid flow rate during storm events. However, not all t�e rain events produce appreciable c�anges in water level or EC. This could suggest t�at parts of t�e rapidly draining fractures are not directly involved in feeding t�e c�annel. It could also be evidence of t�e occurrence of solid precipitation (snow) during t�is period t�at, t�erefore, does not give infiltration.

(10)

The temperature in t�e cave is determined by t�e strong airflow. This influence is particularly evident in t�e sector of t�e cave close to t�e artificial entrance during winter, w�en cold and dry air enters from lower entrances.

During summer, t�e airflow is downward and bot�

t�e stations record t�e temperature of t�e air coming from upper entrances, w�ic� �ere is already in t�ermal equilibrium wit� t�e cave walls. Fig. 3 clearly s�ows t�at w�en t�e external temperature rises systematically above t�e mean cave temperature, t�e two stations soon reac�

almost t�e same temperature, indicating t�at t�e airflow c�anges direction.

It s�ould be noted t�at t�e minimum of temperature at URB-2 is s�ifted later in time wit� respect to sta-

tion ARPAT-2. This be�avior could be explained in t�is way. we �ypot�esize t�at in April t�ere is a downward flow of air, w�ic� runs in galleries wit� flowing waters before arriving in t�e Galleria delle Stalattiti. In t�is period, t�e infiltration water still derives from snow melting and so is relatively colder t�an during t�e summer season. Conversely, t�e Galleria del Venerdì, w�ere t�e station ARPAT-2 is placed, is a dry passage, and in t�is situation t�e air tends toward a t�ermal equilibrium wit�

t�e rock (Badino 2004), w�ic� �ere is warmer t�an flowing waters.

Deep sectors of t�e cave, suc� as t�e Galleria delle Stalattiti, are c�aracterized by more stable t�ermal conditions wit� maximum annual oscillations of less t�an 1°C.

The two monitored sectors of t�e cave also display different �ydrodynamic and p�ysicoc�emical c�aracteristics of t�e dripwater. CNR-1 seems to reply rapidly to

Fig. 10: geological sketch of the 800-900 m asl level of Corchia cave (tourist path), showing the hydro-structural position of monitoring sites. Note that stations CNR-2 and URB-2 are below the overturned basement.

events and t�ere is more variation in t�e data for dripwater temperature, conductivity, pH, and δ¹³C t�an at CNR-2. At CNR-1, t�e rock is a well karstified marble wit� persistent interconnected fractures t�at allow ef- ficient flow of infiltration water. The �ig� variability of drip rate suggests t�e system responds quickly to �ydro- static pressure variations related to rainfall events. we measured a low oxygen isotope composition of t�e dripwater in April, in correspondence wit� rainfall events and melting of snow. Indeed, as suggested by Piccini et al. (2008), CNR-1 is fed by waters wit� lower residence time as indicated by few tritium measurements.

Despite t�e small difference between t�e pH of CNR-1 and CNR-2 during t�e monitored period, t�e CNR-1 median value is lower t�an CNR-2 denoting a possible influence of rain- water, as recorded for La Gronda c�annel. Station CNR-2, in Galleria delle Stalattiti, is geometrically overlain at t�e surface by an impermeable basement of p�yllites and metavolcanics (Fig. 10), w�ic� are only su- perficially fractured (Piccini et al. 2008). Furt�ermore, al- teration processes of silicates produce clay minerals t�at prevent t�e vertical infiltration of rec�arging meteoric water. Therefore, we presume t�at water travel is partially developed along t�e contact zone of “Grezzoni” and t�is basement. Suc� structural conditions could explain t�e different be�avior and t�e long residence time of t�e waters in t�e feeding system. In fact, CNR-2 is a �ig�er conductivity site, indicating a �ig�er content of dissolved solids, compared to CNR-1.

TIC is �ig�er for CNR-2 waters (2.76±0.09 mmole) t�an CNR-1 (2.17±0.07 mmole). This result probably indicates t�at t�e infiltrating waters at CNR-2 become en- ric�ed wit� carbonate along t�eir flow pat�. This could be related to t�e �ig�er residence time, as supported by tritium data (Doveri et al. 2005; Piccini et al. 2008), in t�e system and t�e rock type. As illustrated in t�e Fig. 7 t�e alkalinity and Ca at bot� stations are constant. This supports t�e notion t�at precipitation of carbonate does not �appen in t�e one-mont� sample (as empirically ob- served), even if a CO₂ degassing is probably occurring.

Alt�oug� only one drip point was monitored in t�e Galleria delle Stalattiti preliminary data indicated

(11)

CONCLUSION

REFERENCES

Badino, G., 2004: Cave temperatures and global climate c�ange.- International Journal of Speleology, 33, 1/4, 103-114.

Baldini, J.U.L., McDermott, F. & I.J. Fairc�ild, 2006: Spa- tial variability in cave drip water �ydroc�emistry:

Implications for stalagmite paleoclimate records.- C�em. Geol., 235, 390-404.

The present strong airflow influences t�e air temperature in t�e sector of t�e cave close to t�e artificial entrance particularly during winter, w�en cold and dry air enters from lower entrances. Deep sectors of t�e cave, like t�e Galleria delle Stalattiti, are in any case c�aracterized by sufficiently stable t�ermal conditions wit� maximum annual oscillations of less t�an 1°C.

Hydrological and �ydro-geoc�emical data from La Gronda c�annel indicate t�at t�e epikarst above Corc�ia Cave is �ig�ly conductive and give a general idea of t�e speed wit� w�ic� rain enters t�e system. Monitoring

�ig�lig�ts t�e different be�aviour of t�e system at eac�

station investigated, due to position wit� respect to t�e entrances and dept� wit� respect to t�e surface. Hydro- geoc�emical data and drip counts indicate a significant difference between stations CNR-1 and CNR-2, wit�

t�e latter c�aracterised by more stable conditions. The t�icker rock mass and, moreover, t�e peculiar geological conditions of t�e Galleria delle Stalattiti, explain t�ese differences. This is also in agreement wit� isotopic and tritium data. As already stated in PicciniAs already stated in Piccini et al. (2008), t�e isotopic composition of t�e water (δD, δ¹⁸O and tritium content) at t�e CNR-2 drip indicates a stable and well- mixed plumbing system c�aracterized by a relatively long (ca. 50 yr) residence time. This implies t�at only persis-

tent and relatively long-term (decadal?) variations in t�e isotopic composition of local precipitation are capable of forcing a significant c�ange in t�e isotopic composition of dripwaters and, t�erefore, of t�e speleot�em calcite.

The �ydrodynamic and �ydro-p�ysical be�avior of t�e drip at station CNR-2 and of drip sampling in t�e same sector of t�e cave s�ow constant c�emical and p�ysical values, in agreement wit� t�e presence of a regular plumbing system, as suggested by stable isotopes.

Alt�oug� t�e monitoring of t�e cave was largely dis- continuous and only for t�e recent period, data are collected systematically, so t�e data presented �ere corroborate t�e �ypot�eses suggesting Galleria delle Stalattiti as a good example of a “deep” �ypogean system of Fairc�ild et al. (2007). In fact, t�e p�ysical c�emistry (suc� as pH, conductivity and temperature) at CNR-2 suggests a uniform composition of t�e water and, since t�ere is no correlation wit� rainfall events, t�e component from fracture-fed flow seems to be negligible. Moreover, t�e air temperature in t�at sector s�ows oscillation less t�an 0.8°C. This can explain t�e �ig� quality of t�e Galleria delle Stalattiti in recording centennial to millennial climatic c�ange, w�ereas it is probably not suitable for recording environmental c�anges at �ig�er frequency.

t�at most of t�e c�aracteristics of t�e drip are reason- ably constant over a long period. This indicates t�at any concretions originated by t�is kind of flow will mainly reflect long term c�anges in drip conditions over t�e re- c�arge above t�e cave, rat�er t�an c�anges driven by seasonal/yearly variability. This seems to closely matc� t�e

“�ypogean cave” sensu Fairc�ild et al. (2007). It is prob- It is prob- able t�at t�is condition is principally present in t�e Gal- leria delle Stalattiti for its deep position and t�e presence of t�e impermeable basement on top. Ot�erwise, CNR-1

drips s�ow a more (alt�oug� limited) variable be�aviour as an effect of less rock cover and absence of t�e basement. This implies t�at not all t�e deep cave sector can

�ave t�e same be�aviour. Moreover, t�is indicates t�at potentially t�e speleot�ems growing wit�in t�e Galleria delle Stalattiti can �ave a �ig� paleoclimatic signal/noise ratio reflecting long-term, persistent c�anges in climate conditions. In t�is view t�ese speleot�ems are not able to detect s�ort term (yearly/decadal) environmental c�anges.

(12)

Baker, A., Barnes, w.L. & P.L. Smart, 1997: Stalag- mite drip disc�arge and organic matter fluxes in Lower Cave, Bristol.- Hydrol. Proces., 11, 1541-1555.

Baker, A., Genty, D. & I.J. Fairc�ild, 2000: Hydrological c�aracterisation of stalagmite dripwaters at Grotte de Villars, Dordogne, by t�e analysis of inorganic species and luminescent organic matter.- Hydrol.

Eart� System Sci., 4, 439-449.

Brasc�i, S., Del Freo, P. & L. Trevisan, 1986: Ricostruzi- one degli antic�i g�iacciai sulle Alpi Apuane.- Atti Società Toscana di Scienze Naturali, Mem., s. A, 93, 203-219.

Carmignani, L. & R. Kligfield, 1990: Crustal extension in t�e Nort�ern Apennines: t�e transition from compression to extension in t�e Alpi Apuane Core Complex.- Tectonics, 9, 6, 1275-1303.

Collister, C. & D. Mattey, 2008: Controls on water drop volume at speleot�em drip sites: an experimental study.- Journal of Hydrology, 358, 3-4, 259-267.

Craig, H., 1957: Isotope standards for carbon and oxygen and correction factors for mass-spectrometric analysis of carbon-dioxide.- Geoc�im. Cosmoc�im.

Acta, 12, 133-149.

Craig, H., 1961: Isotopic variations in meteoric waters.- Science, 133, 1702-1703.

Doveri, M., Leone, G., Mussi, M. & Zanc�etta, G., 2005:

Composizione isotopica di acque ipogee nell’Antro del Corc�ia (Alpi Apuane, Toscana nord-occiden- tale). – Memorie Istituto Italiano di Speleologia, S.– Memorie Istituto Italiano di Speleologia, S. Memorie Istituto Italiano di Speleologia, S.

II, 18, 119-132.

Dreybrodt, w., Eisenlo�r, L., Madry, B. & B. Ringer, 1997: Precipitation kinetics of calcite in t�e systemPrecipitation kinetics of calcite in t�e system CaCO₃ – H₂O – CO₂: t�e conversion to CO t�e conversion to CO₂ by t�e slow process HH⁺+HCO₃⁻→CO₂+H₂O as a rate lim-as a rate lim- iting step.- Geoc�im. Cosmoc�im. Acta, 60, 3897- 3904.

Drysdale, R.N., Zanc�etta, G., Hellstrom, J.C., Fallick, A.E., Z�ao, J.x., Isola, I. & G. Brusc�i, 2004: Pal- aeoclimatic implications of t�e growt� �istory and stable isotope (δ¹⁸O and δ¹³C) geoc�emistry of a Middle to Late Pleistocene stalagmite from central- western Italy.- Eart� and Planetary Science Letters, 227, 215-229.

Drysdale, R.N., Zanc�etta, G., Hellstrom, J., Fallick, A.

& J.x. Z�ao, 2005: Stalagmite evidence for t�e onset of t�e Last Interglacial in sout�ern Europe at 129

± 1 ka.- Geop�ysical Researc� Letters, 32, L24708, doi:10.1029/ 2005GL024658.

Drysdale, R.N., Zanc�etta, G., Hellstrom, J., Fallick, A., McDonald, J. & I. Cartwrig�t, 2007: Stalagmite evidence for t�e precise timing of Nort� Atlantic cold events during t�e early last glacial.- Geology, 35, 77-80.

Drysdale, R.N., Hellstrom, J. C., Zanc�etta, G., Fallick, A.E., Sánc�ez Gõni, M.F., Couc�oud, I., McDon- ald, J., Maas, R., Lo�mann, G. & I. Isola, 2009: Evi- dence for Obliquity Forcing of Glacial Termination II.- Science, 325, 1527-1531.325, 1527-1531.

Epstein, S. & T.K. Mayeda, 1953: Variations of oxygen-18 of waters from natural sources.- Geoc�im. Cosmo- c�im. Acta, 4, 213-224.

Fairc�ild, I.J., Frisia, S., Borsato, A. & A.F. Toot�, 2007:

Speleot�ems.- In: Nas�, D.J. & S.J. McLaren (eds.) geochemical Sediments and Landscapes. Blackwell, pp. 200-245, Oxford.

Genty, D., Baker, A. & B. Vokal, 2001: Intra- and inter- annual growt� rate of modern stalagmites.- C�em.

Geol., 176, 191-212.

Gran, G., 1952. Determination of t�e equivalence point in potentiometric titrations- Part II.- The Analyst,- The Analyst,The Analyst, 77, 661-671.

Mantelli, F., Piccini, L., Montigiani, A., Cecconi, E. & L.

Lotti, 2005: Geoc�imica delle acque di circolazione nel Complesso Carsico del Monte Corc�ia.- Memo- rie Istituto Italiano di Speleologia, S. II, 18, 177-189.S. II, 18, 177-189.

McDonald, J., Drysdale, R.N., Hill, D., C�isari, R. & H.

wong, 2007: The �ydroc�emical response of cave drip waters to sub-annual and inter-annual climate variability, wombeyan Caves, SE Australia.- C�em.

Geol., 244, 605-623.

Molli, G. & L. Vaselli, 2006: Structures, interference pat- terns, and strain regimes during midcrustal defor- mation in t�e Alpi Apuane (Nort�ern Apennines Italy).- Geological Society of America, Spec. Pub.

414, 79-93.

Musgrove, M. & J.L. Banner, 2004: Controls on t�e spa- tial and temporal variability of vadose dripwater geoc�emistry: Edwards Aquifer, central Texas.-Aquifer, central Texas.- Geoc�im. Cosmoc�im. Acta, 68, 1007-1020.

Piccini, L., 1998: Evolution of karst in t�e Alpi Apuane (Italy): Relations�ips wit� t�e morp�otectonic �istory.- 4t� Int. Conf. on Geomorp�., Bologna, Sep- tember 1997, Suppl. Geografia Fisica Dinamica quaternaria, 3, 4, 21-31.

Piccini, L., Drysdale, R.N. & H. Heijnis, 2003: Karst caves morp�ology and sediments as indicators of t�e uplift �istory in t�e Alpi Apuane (Tuscany, Ita- ly).- quaternary International, 101-102, 219-227.

Piccini, L, Zanc�etta, G., Drysdale, R.N., Hellstrom, J., Isola, I., Fallick, A.E., Leone, G., Doveri, M., Mussi, M., Mantelli, F., Molli, G., Lotti, L., Concioni, A., Regattieri, E., Mecc�eri, M. & L. Vaselli, 2008: The environmental features of t�e Monte Corc�ia cave system (Apuan Alps, central Italy) and t�eir effects on speleot�em growt�.- International Journal of Speleology, 37, 153-172.

(13)

Reed, M.H., 1982: Calculation of multicomponent c�emical equilibria and reaction processes in systems involving minerals, gases, and an aqueous p�ase.- Geoc�imica and Cosmoc�imica Acta, 46, 513-528.

Spötl., C., Fairc�ild, I.J. & A.F. Toot�, 2005: Cave air con- trol on dripwater geoc�emistry, Obir Caves (Aus- tria): implications for speleot�em deposition in dy- namically ventilated caves.- Geoc�im. Cosmoc�im.

Acta, 69, 2451-2468.

Stumm, w. & J.J. Morgan, 1981: Aquatic Chemistry.-- wiley, pp. 211, New York.

Zanc�etta, G., Drysdale, R.N., Hellstrom, J.C., Fallick, A.E., Isola, I., Gagan, M.K. & M.T. Paresc�i, 2007:

En�anced rainfall in t�e western Mediterranean during deposition of sapropel S1: stalagmite evidence from Corc�ia cave (Central Italy).- quaternary Science Reviews, 26, 279-286.