This article was downloaded by: [Mount Allison University 0Libraries]On: 30 April 2013, At: 12:51Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Isotopes in Environmental and HealthStudiesPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/gieh20

Stable isotope study of precipitationand cave drip water in Florida (USA):implications for speleothem-basedpaleoclimate studiesBogdan P. Onac a , Kali Pace-Graczyk b & Viorel Atudirei ca Department of Geology, University of South Florida, 4202 E.Fowler Ave., SCA528, Tampa, FL, 33620, USAb Tetra Tech EM Inc, Minneapolis, MN, USAc Department of Earth and Planetary Sciences, University of NewMexico, Albuquerque, NM, USAPublished online: 28 May 2008.

To cite this article: Bogdan P. Onac , Kali Pace-Graczyk & Viorel Atudirei (2008): Stable isotopestudy of precipitation and cave drip water in Florida (USA): implications for speleothem-basedpaleoclimate studies , Isotopes in Environmental and Health Studies, 44:2, 149-161

To link to this article: http://dx.doi.org/10.1080/10256010802066174

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.

Isotopes in Environmental and Health StudiesVol. 44, No. 2, June 2008, 149–161

Stable isotope study of precipitation and cave drip waterin Florida (USA): implications for speleothem-based

paleoclimate studies†

Bogdan P. Onaca*, Kali Pace-Graczykb and Viorel Atudireic

aDepartment of Geology, University of South Florida, 4202 E. Fowler Ave., Tampa, SCA528, FL 33620,USA; bTetra Tech EM Inc, Minneapolis, MN, USA; cDepartment of Earth and Planetary Sciences,

University of New Mexico, Albuquerque, NM, USA

(Received 14 January 2008; final version received 22 February 2008 )

Stable isotopes of hydrogen and oxygen were used to examine how the isotopic signal of meteoric water ismodified as it travels through soil and epikarst into two caves in Florida. Surface and cave water sampleswere collected every week from February 2006 until March 2007. The isotopic composition of precipitationat the investigated sites is highly variable and shows little seasonal control. The δ18O vs. δ2H plot showsa mixing line having a slope of 5.63, suggesting evaporation effects dominate the isotopic compositionof most rainfall events of less than 8 cm/day, as indicated by their low d-excess values. The δ18O valuesof the drip water show little variability (<0.6‰), which is loosely tied to local variations in the seasonalamount of precipitation. This is only seen during wintertime at the Florida Caverns site.

The lag time of over two months and the lack of any relationship between rainfall amount and theincrease in drip rate indicate a dominance of matrix flow relative to fracture/conduit flow at each site.The long residence time of the vadose seepage waters allows for an effective isotopic homogenisation ofindividual and seasonal rainfall events. We find no correlation between rainfall and drip water δ18O atany site. The isotopic composition of drip water in both caves consistently tends to resemble the amount-weighted monthly mean rainfall input. This implies that the δ18O of speleothems from these two caves inFlorida cannot record seasonal cycle in rainfall δ18O, but are suitable for paleoclimate reconstructions atinter-annual time scales.

Keywords: cave water; hydrogen-2; Florida; natural variations; oxygen-18; paleoclimate; precipitation;rainfall

1. Introduction

The natural isotope abundances of oxygen (δ18O), carbon (δ13C), and hydrogen (δ2H) inspeleothems (e.g., stalagmites and flowstone) are widely considered sensitive to climate changesas their growth mechanisms are linked to earth’s atmosphere and hydrosphere in a number of ways[1–3]. Many of the intimate processes that control the multi-proxy variations in speleothems arenot yet fully understood. When using precipitation-linked stable isotope proxies or models to

*Corresponding author. Email: bonac@cas.usf.edu†Revised version of a paper presented at the 9th. Symposium of the European Society for Isotope Research (ESIR), 23 to28 June 2007, Cluj-Napoca, Romania.

ISSN 1025-6016 print/ISSN 1477-2639 online© 2008 Taylor & FrancisDOI: 10.1080/10256010802066174http://www.informaworld.com

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

150 Bogdan P. Onac et al.

Figure 1. Map of Florida showing location of sampling sites for rainfall and cave drip water.

investigate past environmental change, the understanding of processes by which the isotope sig-nal is transferred from the atmosphere and near-surface environment to the calcite in speleothemsis crucial.

Isotope fractionation induced by evaporation from any water bodies and condensation duringatmospheric transport of water vapour causes spatial and temporal variations in the oxygen-18 anddeuterium composition of precipitation. Thus, the isotopic systematic of modern precipitationsprovide a particular fingerprint of their origin. This fingerprint is fundamental when investigatingthe provenance of cave drip water as it carries valuable information on the storm system dynamicsassociated with the precipitation and the isotopic variability. For the background information onthe relationship between the two isotopes and the generation of the isotopic composition ofprecipitation, the reader is referred to one of the classic papers by Craig [4] and Dansgaard [5]and thorough reviews such as those of Gat [6] and Clark and Fritz [7].

This study presents the initial findings of an ongoing effort to understand the relationshipbetween isotopic composition of precipitations and drip waters in caves in Florida (Figure 1), aregion in which no work of this type has been performed. The Florida Peninsula lies in a keyregion where tropical cyclones greatly affect the weather patterns. Surface and cave water sampleswere collected each week from February 2006 till March 2007 in order to answer the followingquestions: (1) how does the isotopic signal of meteoric water modify as it travels through soil andepikarst down into the cave; (2) is the isotopic signal completely buffered, or are the weekly to sea-sonal variations recognisable, and (3) what are the characteristics of drip water hydrology, if any?

2. Geographic, geologic, and climatic setting

The research sites are located in the karst landscape of central-western and northern Florida(Figure 1). Three caves were originally selected for this study; Legend (LGD), located on theBrooksville Ridge, Jennings, outside Ocala, and a show cave at Florida Caverns State Park (FLC),outside the city of Marianna (Figure 1). All caves investigated develop in the Ocala Limestone.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 151

The lower part of this carbonate unit consists of grainstones to packstones, locally dolomitised;the upper part shows an increased mud content and is quite friable [8, 9]. The Ocala Limestone isone of the most permeable formations within the Upper Floridan Aquifer (UFA). The rocks thatcomprise the UFA are young (Eocene/Oligocene) and have retained much of their depositionalporosity [10] The majority of storage in the UFA occurs within the matrix, of which permeabilityis between 10−11 to 10−13.8 m2 [10–12]. The UFA can be defined as having triple porosity flow.Groundwater flow occurs through primary interconnected pore spaces, secondary fractures aswell as through karst conduits [10,12]. The UFA is unconfined in central Florida, along the OcalaUplift where a significant number of caves cluster. Recharge to the aquifer occurs by diffuse ordiscrete infiltration through sinkholes and highly permeable sandy soils that cover the carbonateunits, allowing for a rather rapid transfer of water into the aquifer, the main source of water forFlorida and South Georgia [13].

The Ocala Limestone is unconformably overlain by the Suwannee Limestone, a bioturbated,cross-bedded subtidal grainstone to packstone [9]. To the North, the Ocala Limestone is overlainby the Hawthorn Formation, a 25–30 m thick Miocene sequence of phosphatic clays, quartz sand,and dolomitic limestones. The Hawthorn Formation effectively confines the UFA [14].

Legend and Jennings caves lie within a karst terrain with many solution closed-basins, wherethe bedrock is at or near the ground surface. The thin soil blanket present in this area is a mixture ofsand, clay, and organic deposits [15]. Florida Caverns is surrounded by areas of higher elevationcharacterised by thicker loamy and clayey sands derived from the erosion of the AppalachianMountains [16]. The higher clay content towards the north significantly affects the recharge ratesof the area [17].

Florida has three climatic zones, each classified as hot-humid regions. During six months ofthe year temperatures can be above 32 ◦C and the relative humidity can be at, or well above 50%.Northern Florida, somewhat cooler because of its latitude, can have a significant number of daysbetween November and March when temperatures are below 10 ◦C [18]. Central Florida however,has a longer period of high-temperature and high-humidity days year around when compared withthe northern part [18]. Florida can receive up to 175 cm of rain per year [19–20]. Central Floridareceives most rainfall during summer while passing weather fronts bring northern Florida themajority of precipitation during winter [21]. During summer, warm air moving landwards fromthe Atlantic Ocean and the Gulf of Mexico causes afternoon thunderheads and storms, the primarysource of summer rainfall.

3. Sample and analytical procedures

Climatic data, rainfall, and cave drip waters were measured weekly at all three locations duringthe study period (February 2006–February 2007). Data collection at Jennings Cave stopped unex-pectedly on 7 October 2006, when all of the in-cave equipment and part of the external instrumentswere found vandalised and/or stolen. Thus, the data set was not considered in this study.

On the surface of each research site, a Campbell Scientific, Inc. CR10X data logger collectedtemperature and precipitation amount every fifteen minutes and summed the data into hourlyreadings. Precipitation collection was done with a log series tipping bucket rain gauge. This gaugecollected precipitation in 0.0254 cm increments before automatically tipping collected water intoa plastic bottle protected from evaporation by a parafilm cap, following standard water isotopecollection protocols [22]. Precipitation samples were collected at the same time as the cave dripwater and both were stored at 4 ◦C until analysed.

As cave drip water collection is not standardised, and collection devices are not available topurchase, drip waters were collected using a homemade funnel and hose set-up. A funnel was

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

152 Bogdan P. Onac et al.

zip-tied to the chosen speleothem; a 3-mm silicon hose was secured to the funnel and fed intoa 1L Nalgene bottle capped with parafilm. The relative humidity in both caves was in excess of95%, thus evaporation effects were considered minimal.

Drip rates within the cave were collected using a Stalagmate integrated drip counter andlogger produced by Driptych. The Stalagmate calculates drip frequency by counting the totalnumber of drips landing on a microphone located on the top of the data logger. The microphonesensor has an adjustable sensitivity, which can be preset to record drips falling from as low as50 cm. The microphone is tuned to exclude any extraneous noise and spurious events. The datalogger was placed so that only the drips from a sole stalactite would land on the microphone. TheStalagmate was so designed that it will only count and sum drips each 15 min (time interval isadjustable) during the sample collection. This allows one to estimate (based on changes in driprate) the time needed for water to travel from surface down into the cave.

Gemini TinyTag 2 Plus (TGP-4500) data loggers collected temperature in each cave every 60 s.These were then averaged to give hourly readings throughout the sampling period.All downloadeddata (i.e., temperature and drip rate) were converted to daily and monthly values to be used in thefinal data interpretations.

Determination of δ2H and δ18O for drip water and precipitation samples was performed inthe Stable Isotope Laboratory at the University of New Mexico in Albuquerque (New Mexico).A Finnigan-MAT Delta XP Plus isotope ratio mass spectrometer, automated by the GasBenchII and GC PAL was used for all analyses. Epstein and Mayeda’s [23] method of CO2-waterheadspace equilibration technique was used to analyse oxygen samples while hydrogen sampleswere reacted with high temperature chromium via a H-device. Results were reported in δ unitsvs. the international Vienna-standard mean ocean water (SMOW) [4]. Analytical precisions (2σ )were better than ±1‰ and ±0.1‰ for δ2H and δ18O, respectively.

Given the range in number of water samples collected from precipitations at each cave site andthe small number of samples in some months, the volume-weighted average annual δ values werepreferred over the simple arithmetic average of the same values. The δ values were weighted bymultiplying the isotopic composition of each individual water sample by the discharge measuredat the collection date, and dividing by the total rainfall amount for the dates sampled [19, 20] Theresults indicate that isotope means are slightly lower when weighted with the precipitation amount.

4. Results and Discussion

4.1. Stable isotopic composition of precipitation

The precipitation samples collected above the two caves have values ranging from 0.1 to −5.4‰for δ18O and from 5.3 to −32.3‰ for δ2H, whereas the deuterium-excess (d) is lower than 15‰(Table 1). These values may indicate the existence of cloud fronts coming from both the Gulf ofMexico and the Atlantic. Mean δ18O and δ2H values at LGD are −2.6‰ and −11.9‰, whereasat FLC are −2.9‰ and −13.2‰ respectively.

Scatter plots of δ18O and δ2H (unweighted monthly average values) for each cave site areshown in Figure 2. The local meteoric water lines (LMWLs) from LGD and FLC are not statis-tically different. Thus, when combining the data from both sites the resulting LMWL is definedas δ2H = 5.63δ18O + 3.24 (R2 > 0.96). The slope and intercept are different from the globalmeteoric water line (GMWL) reported by Rozanski et al. (Figure 2) [24], but similar to otherregions in the southern US, where most summer rains originate from the Gulf of Mexico [19, 25].

LMWL slopes less than eight would normally indicate evaporative conditions. Nevertheless,the isotopic values of less-intense rain showers like those typical for La Niña years in Florida plotalong the evaporation lines that have d-excess lower than 15‰ [19]; such deviations were ascribed

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 153

Table 1. Isotopic composition of precipitation (pr) and drip water (dw), air temperature, and amount of rainfall and dripat Legend and Florida Caverns (averages of monthly weighted data 2006/2007).

δ18O pr δ2H pr d-excess δ18O dw δ2H dw Rainfall Drips Air temp.Month (‰) (‰) (‰) (‰) (‰) (cm) (month) (◦C)

Florida CavernsFebruary 2006 −1.20 −2.0 7.61 −3.90 −19.5 2.11 − 12.95March −2.62 −12.0 8.89 −3.93 −19.8 1.34 − 15.50April −2.02 −7.6 8.55 −3.87 −19.1 2.92 − 20.45May −3.39 −15.5 11.56 −3.78 −17.9 11.76 435 22.07June −4.19 −19.2 14.34 − − 1.77 32 25.08July −2.01 −8.5 31.53 − − 12.85 2 26.28August −3.02 −13.1 11.02 − − 11.76 − 26.12September −4.86 −25.1 13.79 − − 2.26 − 23.09October −3.91 −18.5 12.82 − − 6.32 1 18.53November −2.41 −10.5 8.83 − − 9.12 2 13.22December −2.93 −12.6 10.82 − − 13.94 − 12.31January 2007 −2.98 −13.5 10.36 −3.57 −15.6 13.26 − 11.12February −3.92 −18.7 12.60 − − 6.45 1 9.40

Legend CaveFebruary 2006 − − − −3.1 −15.1 − − −March −3.10 −15.1 9.79 −2.70 −12.5 1.65 − 16.85April −1.81 −6.2 8.28 −3.10 −15.0 2.94 29927 21.00May −2.66 −11.7 9.58 −3.02 −14.6 5.18 76067 22.82June −3.01 −15.5 8.63 −3.24 −14.7 48.38 53987 25.47July −3.22 −14.1 11.62 −3.11 −15.2 19.23 51391 26.52August −2.38 −10.9 8.15 −2.99 −14.2 22.51 139330 26.44September −4.33 −21.2 13.48 −3.23 −15.4 10.90 47579 24.63October −2.15 −7.2 10.00 −3.07 −14.1 7.924 1733 24.63November −2.84 −13.6 9.09 −3.10 −15.3 2.57 2497 16.68December −1.96 −8.5 7.17 −3.25 −15.6 10.77 1951 17.63January 2007 −1.40 −7.1 4.11 −3.12 −15.9 7.34 2354 15.69February −2.18 −8.4 9 −2.92 −15.5 3.12 − 12.43

Figure 2. The δ18O–δ2H composition of precipitation, February 2006 through February 2007, compared with theGMWL.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

154 Bogdan P. Onac et al.

Figure 3. Relationship between monthly mean δ18O in precipitation and air temperature. There is a broad and poorpositive correlation in the winter rainfall (see text for details).

to evaporation of droplets beneath the cloud [6]. Hence, it is unclear whether the precipitations atinvestigated sites have undergone evaporation during travel or during rainfall.

As expected for tropical and sub-tropical sites where mean annual temperatures are above15 ◦C, no relationship between δ18O values in precipitation and average monthly temperature wasfound [26]. On the seasonal basis, however, there is a very poor positive correlation (R2 = 0.34),but only at FLC, and only for the isotopic composition measured in rainfalls from November toMarch (Figure 3). The slope of δ18O – T for the winter period at FLC is below the global average[27], suggesting temperature differences between winter and summer being more significant atthis location compared with LGD.

On the 2006–2007 year basis, there is a poor correlation between rainfall amount and oxygenisotope values at the two sites (Figure 4). During certain months, rainfall δ18O values are loweralthough the amount of precipitation decreased. This is not surprising for a region like Florida,since the amount effect is based at least partly on rainfall intensity rather than on the overallamount [5]. On seasonal basis, there is a less agreement, except during the months of Marchthrough August at LGD, which show a poor correlation in the typical amount-effect mode, i.e.,the tendency of higher amounts of rainfall to be more isotopically depleted. This fits the conceptthat summer rainfall in central Florida is derived from convective events rather than weather frontsthat bring winter rainfall.

While for the most sites δ18O vs. δ2H show good correlations for rainfall, some variation canoccur as a result of atmospheric conditions. A useful diagnostic tool of this is the deuterium excessor d-value [7, 27], defined as

d(‰) = δ2H − 8δ18O.

The d-value varies with the relative humidity and wind speed of the air masses at their oceanicorigin, sea surface temperature during primary evaporation, and kinetic isotope effects duringevaporation. Typically, the d-excess of rain samples in temperate climates averages +10‰.Because of the link to humidity, d-excess is sensitive to evaporation processes, including whethersummer or winter rainfall dominates recharge.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 155

Figure 4. Temporal evolution of the average δ18O vs. mean monthly rainfall amount during the study period.

The d-excess at our sites varies in the range of +4 to +15‰, overlapping the spatial distributionpresented by Kendall and Coplen [18] for Florida. There is no correlation between d−excessand mean annual temperature or with precipitation amount. We found that the lowest values ofour record (<+10‰) occurred from October to April during rainfalls with less than 8 cm/day(Figure 5). These values are likely associated with samples affected by evaporation. High d-excess values (>+10‰) are generally seen during summer rainfalls, indicating more evaporatedmoisture being added to the atmosphere. No correlation between d-excess and rainfall intensitywas found.

4.2. Drip water hydrology and stable isotope composition

The characteristics of water percolation through the soil and epikarst, and drip hydrology arecrucial for understanding the environmental factors controlling the cave drip water isotopic com-position and the geochemical signals preserved in speleothems [28]. In order to evaluate theeffect of rainfall recharging and mixing within the vadose zone, one stalactite in each cave wasmonitored weekly. Both stalactites had an average drip rate of less than 4 drips per minute (dpm).

The monitored drip site in LGD was located on the ceiling of the main chamber, which isabout 6 m below the surface. The stalactite drip rate varied between 1 to 16 dpm and the drippingcontinued all year long, despite a soil moisture deficit during winter 2005 and spring 2006. Thishydrologic behaviour suggests that there must be a significant storage flow component to this dripsuch that the drip is maintained all year.

A plot of rainfall amount vs. drip rate shows no discernable correlation (Figure 6). Comparisonof the timing of the two most apparent drip rate increases (10 August 2006 and 19 August 2006)and the prevailing climate (Figure 6, inset) demonstrates that both events occurred ∼63 days aftertwo stormy days (3 June 2006 and 21 June 2006) that delivered up to 22 cm of water per day abovethe cave. These were all associated with disturbed weather originating in the Gulf of Mexico. Nodistinct drip rate peaks were generated by rainfall events with less than 8 cm/day. The mechanism

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

156 Bogdan P. Onac et al.

Figure 5. Deuterium-excess (d) values of precipitation events at Legend and Florida Caverns vs. rainfall amount.Rainfalls <8 cm show lower d values.

Figure 6. Plot of rainfall amount above the two cave sites vs. drip rate. Note the time lag between the events at LegendCave and the lack of signal at Florida Caverns (drip values at LGD are divided by 1000).

explaining this behaviour of drip rate might be coupled with rainwater being gradually storedin the unsaturated zone, from where an increase in hydraulic head during summer causes latedischarge due to piston flow effect. This also suggests that there is no fissure or karst conduit flowcomponent above LGD, otherwise the transfer of the rainwater down into the cave would happenfaster.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 157

Figure 7. Delta-plot showing monthly cave drip water isotopic composition in comparison with the GMWL.

At FLC, the drip water sample location was on a heavily decorated chamber, off the show-cavepath, about 10 m below the surface. Despite an averaged 7.37 cm of precipitation per month,beginning with July 2006 the drip rate decreased down to 1–2 drips/month, ceasing completelyfor four months (August, September, December 2006, and January 2007). During this time therewere no active dripping stalactites throughout the cave. The available data indicate this dripsite has very little variability with any seasonal trend, lagging any increase in hydrologicallyeffective precipitation by ∼8 months. This suggests the existence of a storage zone around thecave (characterised by matrix flow) that holds precipitation water before it drips into the cave.The clayey sediments in the soil blanket may also affect the recharge.

The drip water in LGD exhibits a range of δ18O from −4.4‰ to −1.9‰ (n = 52), which isrelatively large compared with FLC (−4.0 to −3.6‰, n = 10). The weighted-average drip waterδ18O value is −3.1‰ in LGD and −3.8‰ in FLC, slightly lower than the weighted-averagemonthly rainwater δ18O value for each individual site. The δ2H of drips collected in the twocaves fall in a narrow range between −16 and −20‰ in FLC and −12 and −16‰ in LGD. Thenarrow range observed for cave drip water composition reflects attenuation of the seasonal rainfallisotopic variation due to mixing of recharge water from different events in the epikarst.

The cave drip water isotope values fall in two distinct fields along the GMWL (Figure 7). Thefive available samples (all collected over the winter period) from FLC plot far to the lower-leftand above the GMWL (δ2H = 11.74δ18O + 26.38; R2 > 0.999; n = 5). Active air circulationdriven by differences in air temperature between the cave and outside during the winter increasesthe evaporation rate in the cave atmosphere. Thus, the enrichment of drips probably occurs whilethe water is temporarily stored at the tip of the stalactite. The LMWL derived from the isotopiccomposition of drip water from LGD (δ2H = 4.36δ18O − 1.43; R2 > 0.55; n = 13) has a lowerslope compared with the rainfall-based LMWL. This may reveal a significant contribution ofrainfall evaporation near the surface or in the soil.

Some of the previous studies showed that the isotopic values of drip water in caves representthe mean annual value of rainfall in the area in which the cave lies [29–31]. This implies thatthe water stored in the aquifer has undergone complete homogenisation. A number of studies,however, report no effective homogenisation as rainfall water percolates through the epikarstand, therefore, seasonal cycles are recorded in speleothems and karst springs [32–34]. Other

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

158 Bogdan P. Onac et al.

investigations found differences between the δ18O values of rainfall and drip water collected fromslow or fast dripping stalactites at different cave locations or even within the same cave [35–38].Such stalagmites growing under fast-dripping stalactites are likely to record paleo-seasonal cyclesin rainfall [39–40].

We find no correlation between isotopic values in precipitation and cave drip water at any of thesites (Figure 8). This suggests that within the epikarst, an effective mixing of recharge water gener-ated by different rainfall events has occurred. However, a weak but somehow significant correlation

Figure 8. Relationship between δ18O values in precipitation and cave drip water at each site. Note the lack of correlationbetween the two data sets suggesting an effective mixing of water in the epikarst.

Figure 9. Plot of drip water δ18O vs. drip rate at LGD.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 159

(R2 = 0.21) exists between drip rate and δ18O values in drip water from LGD (Figure 9), probablyreflecting a coherent response to inter-annual precipitation variability.

5. Concluding remarks

Rainfall and cave drip water isotope data for two locations in Florida have been investigated.One year’s worth of weekly data from LGD and FLC exhibited large fluctuations in precipitationisotope value over short periods of time. Correlations with meteorological parameters are notstraightforward to interpret, but a detailed study at the synoptic scale (beyond the scope of thispaper) remains to be made.

Our work provides a first step towards an understanding of how rainfall-cave drip water δ18Orelationship at two locations in Florida can improve interpreting δ18O of speleothems. The isotopiccomposition of the percolating waters intercepted in Legend and Florida Caverns shows variousrelationships with the rainfall regime (intensity, duration), surface temperature, evaporation, andthe mode of water moving within the soil and epikarst. Accordingly, the main conclusions are asfollows.

The isotopic composition of precipitation at the two investigated sites is highly variable andhas little seasonal control. Different pathways of the rain-bringing air masses might at least inpart explain this variability. The δ18O vs. δ2H plot shows a mixing line having a slope of 5.63.Fractionation by evaporation dominates the isotopic composition of most rainfall events with lessthan 8 cm of rain, as indicated by their low d-excess values [6].

The relationship between rainfall amount and drip rate suggests that there is a time lag oftwo months or more between intensive storm events (>8 cm/day) and an increase in drip rate.The long residence time of the vadose seepage waters indicate that matrix flow dominates overfracture/conduit flow within the epikarst above LGD. At FLC, the long residence time maypossibly reflect the presence of thick clayey sediments of the Hawthorn Group that cause a delayin recharge.

Our results indicate that there is no relationship between rainfall and drip water δ18O variabilityat either site. With homogenisation occurring along the flow path and residence times on the orderof months, the very small variations (<6‰) observed in the isotopic composition of drip watercannot be related to seasonal variations in the isotopic composition of local precipitation. Theisotopic composition of drip water in both caves consistently tends to resemble the amount-weighted monthly mean rainfall input. This is further confirmed by the lack of correspondencebetween mean daily drip rate and rain events, which indicates that the isotopic signature of anygiven rainfall event is highly dampened in the epikarst. Therefore, slow-growing stalagmites inany of these caves would likely record changes in annual average of rainfall δ18O.

Drip water δ18O variability is to some extent controlled by rainfall δ18O variability, which isloosely tied to the local precipitation amount on seasonal time scale (at FLC only during thewinter, and only based on very few data). Because the seasonal isotopic amplitudes in Florida arelow, a large seasonal fluctuation in precipitation amount may result in undetectable inter-annualisotopic signals. Thus, significant changes in δ18O along the growth axis of a speleothem cannotbe a priori attributed to changes in temperature.

Although the drip waters at each site appear to reflect the amount-weighted annual isotopiccomposition of the precipitation waters outside the cave, the climate–rainfall (amount, δ18O, δ2H)relationships are not always simple to determine. The overall conclusion for the interpretation ofδ18O and δ2H in speleothems is that a thorough understanding of the detailed processes controllingthe modern variability in stable isotopic composition of precipitation and the hydrological factorsat the appropriate scale(s) are necessary to be fully confident of the validity of the paleoclimaticinterpretations.

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

160 Bogdan P. Onac et al.

Acknowledgements

This study was supported by grants from the University of South Florida (BPO), Geological Society of America GraduateResearch Grant, National Speleological Society Research Grant, University of South Florida, and Florida Studies CenterResearch Grant (KPG). We wish to thank Southwest Florida Water Management District (Steve Saxton) for providingthe equipment for meteorological logging, Colleen Werner, Mike Gordon, and Florida Caverns State Park for providingpermits to enter and collect samples from Legend, Jennings, and Florida Caverns, respectively. The authors are indebtedto T. Turner, R. Brooks, J. Hagan, M. Gordon, L. Florea, L. Soto, B. Fratesi, A. Persoiu, and G. Hunt for field assistance.The isotope database for each research site is available at: http://www.nrel.colostate.edu/projects/usnip/.

References

[1] J.A. Dorale, R.L. Edwards, and B.P. Onac, Stable isotopes as environmental indicators in speleothems, in KarstProcesses and the Carbon Cycle, edited by D. Yuan and C. Zhang (Geological Publishing House, Beijing, 2002),pp. 107–120.

[2] F. McDermott, H.P. Schwarcz, and P.J. Rowe, Isotopes in speleothems, in Isotopes in Palaeoenvironmental Research,edited by M.J. Leng (Springer, Berlin, 2005), pp. 185–225.

[3] I.J. Fairchild, C.L. Smith, A. Baker, L. Fuller, C. Spotl, D. Mattey, and F. McDermott, Modification and preservationof environmental signals in speleothems, Earth Sci. Rev. 75, 105 (2006).

[4] H. Craig, Isotopic variations in meteoric waters, Science 133, 1702 (1961).[5] W. Dansgaard, Stable isotopes in precipitation, Tellus 16, 436 (1964).[6] J.R. Gat, Oxygen and hydrogen isotopes in the hydrologic cycle, Annu. Rev. Earth Planet. Sci. 24, 225 (1996).[7] I. Clark and P. Fritz, Environmental Isotopes in Hydrogeology (Lewis Publishers, Boca Raton, 1997).[8] R.E. Copeland, T.M. Scott, J.M. Lloyd, and G.L. Maddox, Florida’s Ground Water Quality Monitoring Program:

Hydrogeologic Framework (Spec. Publ.) (Florida Geological Survey, Tallahassee, 1991), Vol. 32.[9] A.F. Randazzo, The sedimentary platform of Florida: Mesozoic to Cenozoic, in The Geology of Florida, edited by

A.F. Randazzo and D.S. Jones (University Press of Florida, Gainesville, 1997), pp. 39–56.[10] D.A. Budd and H.L. Vacher, Matrix permeability of the confined Floridan Aquifer, Florida, USA, Hydrogeol. J. 12,

531 (2004).[11] S.R.H. Worthington, D.C. Ford, and P.A. Beddows, Porosity and permeability enhancement in unconfined carbonate

aquifers as a resolut of solution, in Speleogenesis. Evolution of Karst Aquifers, edited by A. Klimchouk, D.C. Ford,A. Palmer, and W. Dreybrodt (National Speleological Society, Huntsville, 2000), pp. 463–472.

[12] L.J. Florea and H.L.Vacher, Springflow hydrographs: eogenetic vs. telogenetic karst, Ground Water 44, 352 (2006).[13] J.A. Miller, Hydrogeology of Florida, in The Geology of Florida, edited byA.F. Randazzo and D.S. Jones (University

Press of Florida, Gainesville, 1997), pp. 69–88.[14] T.M. Scott, Miocene to Holocene history of Florida, in The Geology of Florida, edited by A.F. Randazzo and D.S.

Jones (University Press of Florida, Gainesville, 1997), pp. 57–67.[15] R.L. Myers and J.J. Ewel, Ecosystems of Florida (University of Central Florida Press, Orlando, 1990).[16] W. Schmidt, Geomorphology and physiography of Florida, in The Geology of Florida, edited by A.F. Randazzo and

D.S. Jones (University Press of Florida, Gainesville, 1997), pp. 1–12.[17] B.G. Katz, J.S. Catches, T.D. Bullen, and R.L. Michel, Changes in the isotopic and chemical composition of

groundwater resulting from a recharge pulse from a sinking stream, J. Hydrol. 211, 178 (1998).[18] N.O.A.A. (2007) Available online at: http://www7.ncdc.noaa.gov/IPS/CDPubs?action=getpublication.[19] C. Kendall and T.B. Coplen, Distribution of oxygen-18 and deuterium in river waters across the United States,

Hydrol. Process. 15, 1363 (2001).[20] R.W. Vachon, J.W.C. White, E. Gutmann, and J.M. Welker, Amount-weighted annual isotopic (δ 18O) values affected

by the seasonality of precipitation: a sensitivity study, Geophys. Res. Lett. 34, L21707 (2007).[21] J.B. Martin and S.L. Gordon, Groundwater flow and contaminant transport in carbonate aquifers, in Surface and

Groundwater Mixing, Flow Paths, and Temporal Variations in the Chemical Composition of Karst Spring, edited byI.D. Sasowsky and C.M. Wicks (Taylor & Francis, Rotterdam, 2000), pp. 65–92.

[22] I. Friedman, Deuterium content of natural waters and other substances, Geochim. Cosmochim. Acta 4, 89 (1953).[23] S. Epstein and T. Mayeda, Variation of O18 content of waters from natural sources, Geochim. Cosmochim. Acta 4,

213 (1953).[24] K. Rozanski, L. Araguas-Araguas, and R. Gonfiantini, Isotopic patterns in modern global precipitation, in Climate

Change in Continental Isotopic Record, edited by P.K. Swart, K.C. Lohmann, and J. McKenzie (AmericanGeophysical Union, Geophysical Monograph Series, 1993), Vol. 78, pp. 1–36.

[25] J.M. Welker, Isotopic (δ 18O) characteristics of weekly precipitation collected across the USA: an initial analysiswith application to water source studies, Hydrol. Process. 14, 1449 (2000).

[26] J. Jouzel, K. Froehlich, and U. Schotterer, Deuterium and oxygen-18 in present-day precipitation: data and modelling,Hydrol. Sci. J. 42, 747 (1997).

[27] Z. Sharp, Principles of Stable Isotope Geochemistry (Pearson Prentice Hall, Upper Saddle River, 2007).[28] A. Baker and C. Brunsdon, Non-linearities in drip water hydrology: an example from Stump Cross Caverns,Yorkshire,

J. Hydrol. 277, 151 (2003).

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013

Isotopes in Environmental and Health Studies 161

[29] R.S. Harmon and R.L. Curl, Preliminary results on growth rate and paleoclimate studies of a stalagmite from OgleCave, New Mexico, Natl. Speleol. Soc. Bull. 40, 25 (1978).

[30] C.J. Yonge, D.C. Ford, J. Gray, and H.P. Schwarcz, Stable isotope studies of cave seepage water, Chem. Geol. 58,97 (1985).

[31] E. Caballero, C. Jimenez de Cisneros, and E. Reyes, A stable isotope study of cave seepage waters, Appl. Geochem.11, 583 (1996).

[32] B. Li, D.Yuan, J. Qin, Y. Lin, and M. Zhang, Oxygen and carbon isotopic characteristics of rainwater, dripwater andpresent speleothems in a cave in Guilin area, and their environmental meanings, Sci. China 43, 277 (2000).

[33] L. Aquilina, B. Ladouche, and N. Dorflinger, Recharge processes in karstic systems investigated through thecorrelation of chemical and isotopic composition of rain and spring-waters, Appl. Geochem. 20, 2189 (2005).

[34] F.W. Cruz, Jr., I. Karmann, O. Viana, , S.J. Burns, J.A. Ferrari, M. Vuille, A.N. Sial, and M.Z. Moreira, Stable isotopestudy of cave percolation waters in subtropical Brazil: implications for paleoclimate inferences from speleothems,Chem. Geol. 220, 245 (2005).

[35] M. Bar-Matthews, A. Ayalon, A. Matthews, E. Sass, and L. Halicz, Carbon and oxygen isotope study of theactivewater-carbonate system in a karstic Mediterranean cave: implications for paleoclimate research in semiaridregions, Geochim. Cosmochim. Acta 60, 337 (1996).

[36] J.U.L. Baldini, F. McDermott,A. Baker, L.M. Baldini, D.P. Mattey, and L.B. Railsback, Biomass effects on stalagmitegrowth and isotope rarios: a 20th century analogue from Wiltshire, England. Earth Planet. Sci. Lett. 240, 486 (2005).

[37] P.E. van Beynen and P. Febbroriello, Seasonal isotopic variability of precipitation and cave drip water at Indian OvenCave, NewYork, Hydrol. Process. 20, 1793 (2006).

[38] K.M. Cobb, J.F. Adkins, J.W. Partin, and B. Clark, Regional-scale climate influences on temporal variations ofrainwater and cave dripwater oxygen isotopes in northern Borneo, Earth Planet. Sci. Lett. 263, 207–220 (2007).

[39] P.C. Treble, J. Chapell, M.K. Gagan, K.D. McKeegan, and T.M. Harrison, In situ measurements of seasonal δ 18Ovariations and analysis of isotopic trends in a modern speleothem from southwest Australia, Earth Planet. Sci. Lett.233, 17 (2005).

[40] K.R. Johnson, B. Lynn Ingram, W.D. Sharp, and P. Zhang, East Asian summer monsoon variability during marineisotope Stage 5 based on speleothem δ 18O records from Wanxiang Cave, central China, Palaeogeo. Palaeoclim.Palaeoecol. 236, 5 (2006).

Dow

nloa

ded

by [

Mou

nt A

lliso

n U

nive

rsity

0L

ibra

ries

] at

12:

51 3

0 A

pril

2013