Stable isotopes and trace elements of drip waters at DeSoto Caverns during rainfall-contrasting years

Accepted Manuscript Stable isotopes and trace elements of drip waters at DeSoto Caverns during rainfall-contrasting years

Rajesh Dhungana, Paul Aharon PII: DOI: Reference:

S0009-2541(18)30548-5 https://doi.org/10.1016/j.chemgeo.2018.11.002 CHEMGE 18964

To appear in:

Chemical Geology

Received date: Revised date: Accepted date:

29 March 2018 27 October 2018 10 November 2018

Please cite this article as: Rajesh Dhungana, Paul Aharon , Stable isotopes and trace elements of drip waters at DeSoto Caverns during rainfall-contrasting years. Chemge (2018), https://doi.org/10.1016/j.chemgeo.2018.11.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1 Stable Isotopes and Trace Elements of Drip Waters at DeSoto Caverns During RainfallContrasting Years

35487, USA

CE

PT

ED

M

AN

US

Corresponding author ([email protected])

AC

2

IP

Department of Geological Sciences, Box 870338, University of Alabama, Tuscaloosa, AL

CR

1

Rajesh Dhungana and 1,2Paul Aharon

T

1

ACCEPTED MANUSCRIPT 2 Abstract A monitoring study at DeSoto Caverns during two years (2012-2013) of rainfall-contrasting variability presents the opportunity to test the response of the hydroclimate proxies in drips and active speleothems to forcing factors on intrannual and interannual time scales.

T

The weighted monthly mean rainwater δ18O and δ2H range from -1.2 to -6.4 (‰ V-SMOW)

IP

and -4 to -41.6 (‰ V-SMOW), respectively, and show modest interannual variation. D-excess

CR

values exhibit a large intrannual contrast suggesting a primary control by sub-cloud evaporation processes. Coeval drip-water δ18O and δ2H vary from -3.1 to - 5.3 (‰ V-SMOW) and -9.9 to

US

-30.5 (‰ V-SMOW), respectively, and exhibit interannual negative trends from the 2012

AN

dry/warm year to the 2013 relatively wet/cool year. Substantial attenuation of drip-water isotope

evaporated-water in the epikarst zone.

M

amplitudes, relative to its rainwater source, is likely caused by mixing of fresh with residual

ED

Drip-water Ca, Mg, Sr and Mg/Ca and Sr/Ca ratios exhibit an inverse relation with respect to the contrasting hydroclimate years such that lower values and higher ratios occur during the

PT

dry/warm year and higher values and lower ratios occur during the wet/cool year. We assert that

CE

interannual rainfall variability exerts a dominant control on the elemental concentrations and their ratios of the drips through changes of biomass productivity in the soils overlying the cave,

AC

and prior aragonite precipitation in the epikarst. The distribution coefficients of Mg (DMg = 3.49x10-3 ±1.06x10-3) and Sr (DSr = 1.12 ± 0.041) between drips and aragonite speleothems estimated in this study are in broad agreement with aragonite-solute experimental values. Coeval changes of trace elements and δ18O in response to interannual rainfall variability confirm their usefulness to better constrain the controlling hydroclimate drivers.

ACCEPTED MANUSCRIPT 3 Keywords Southeast USA; DeSoto Caverns; rainfall and drip-waters; aragonite speleothems; trace- element partition coefficients; oxygen and hydrogen isotopes.

T

1. Introduction

IP

Speleothem-based climate proxies such as δ18O and trace-element ratios (e.g., Mg/Ca and Sr/

CR

Ca) have been successfully used to understand past climate changes on account of their potential to provide continuous records of past rainfall, temperature variability and atmospheric circulation

US

patterns on various timescales (Burns et al., 2001; Cruz et al., 2007; Wang et al., 2008; Mattey et

AN

al., 2008; Fairchild and Treble, 2009; Wong et al, 2011). Coeval occurrence of multiple climate factors impacting the geochemical proxies, however, often makes their interpretation tenuous.

M

For example, observed δ18O variability in speleothems can be caused by changes in rainfall

ED

amount, temperature, moisture sources and disequilibrium isotope effects (Lachniet, 2009; Feng et al., 2013). Similarly, variations of trace-element ratios in speleothems can be caused by soil

PT

and bedrock trace-element compositional variability, changes in water-rock interaction and

CE

groundwater residence time, prior calcite or aragonite precipitation (PCP or PAP, respectively) in the epikarst and drip water flow-routing shifts (Fairchild et al., 2000; McDonald et al., 2007;

AC

Karmann et al., 2007; Riechelmann et al., 2014). In contrast to calcite speleothems, the traceelement partitioning into their aragonite counterparts is poorly constrained (Wassenburg et al., 2016). Whereas δ18O composition of drip-waters and speleothems in tropical and subtropical caves are generally interpreted in terms of rainfall amount variability at a roughly constant mean annual temperature and >95% relative humidity (Burns et al., 2002; Fleitmann et al., 2003;

ACCEPTED MANUSCRIPT 4 Tremaine et al., 201l), the paleoclimate interpretation of cave drip-water Mg/Ca and Sr/ Ca ratios variability is depended on the cave specific processes. For example, Mattey et al. (2010) and Wong et al. (2011) have argued that seasonal variability is caused by cave ventilation shifts whereas Johnson et al. (2006) and Oster et al. (2012) suggested that seasonal rainfall shift is the

T

dominant factor controlling the observed drip-water trace-element ratios variability.

IP

Here we report the results of a two-year monitoring study carried out monthly from the thinly

CR

(~ 10 m) capped large front chamber at DeSoto Caverns (Fig. 1) in order to improve our understanding of the complexities associated with seasonal and interannual variability of drip-

US

water geochemical proxies. Drip-water stable isotope measurements were previously conducted

AN

over a three-year period (June 2005 to May 2008) from the thickly (~ 40 m) capped small backchamber of the cave (Lambert and Aharon, 2010). This study aims to investigate the: (i)

M

relationship between monthly rainfall, drip- rates and estimated water budget; (ii) spatial and

ED

temporal drip-water oxygen and hydrogen isotope variability, and (iii) spatial and temporal variations in drip-water trace elements (Mg and Sr) and their ratios to the major Ca cation

PT

(Mg/Ca and Sr/ Ca). Finally, we discuss the implications of geochemical proxies variability in

CE

drip-waters concerning the paleohydrology archived in the DeSoto Caverns stalagmites. 2. Study site and regional climate

AC

Cave drip-waters were collected from DeSoto Caverns (86°16'36" W, 33°18'26" N) located on the outskirts of Childersburg, AL at the foothills of the Appalachian Mountains. The cave, situated at an elevation of 170 m, lies within the Upper Ordovician dolomite karst area. It is approximately 150 m long and consists of two contiguous segments: (i) a large, 70 m long and 50 m wide, front chamber with ~10 m thick overlying bedrock, and (ii) small chambers in the back of the cave with relatively thicker (~30-40 m) overlying bedrock (Fig. 1). Both inactive and

ACCEPTED MANUSCRIPT 5 active speleothems found within the cave are deposited as aragonite (Lambert and Aharon, 2010; Aharon and Dhungana, 2017). The climate of the study area is humid subtropical and rainfall occurs throughout the year (Baigorria et al., 2007). The average monthly air temperatures range from ~7oC in January to

T

~27oC in July while average monthly rainfall ranges from ~80 mm in October to ~150 mm in

IP

March. The area experiences a mean annual air-temperature of 17.2 oC and an average annual

CR

rainfall of 1417 mm, based on 1958-2004 data (Lambert and Aharon, 2010). Rainfall is usually well distributed throughout the year with two distinct modes. In the winter the collision of

US

opposing cold and warm air masses often results in rain-producing storm systems that are carried

AN

west to east by the polar jet stream (Baigorria et al., 2007). In contrast, summer rainfall is typically derived by convection style thunderstorms whose frequency is influenced by the east-

M

west position of the Bermuda High pressure cell (Lambert and Aharon, 2010; Ortegren et al.,

ED

2011). 3. Methods

PT

The cave was monitored monthly for a two-years period (Jan. 2012 to Dec. 2013), including

CE

sampling of drip-waters for geochemical analysis, with the goal of unraveling the dominant factors controlling the geochemistry of drip waters and thereby of the speleothems.

AC

3.1. Field Methods

Weekly- integrated rainwater samples were collected during the study period from the raingauge station established at the University of Alabama following the method described by Lambert and Aharon (2010). Cave drip-waters were collected from four locations in the front chamber of the cave where the cap rock is relatively thin and the drip falling distance is ~25 m (Fig. 1). The aerial distance between the rain-gauge and the cave sites is ~113 km. Among the

ACCEPTED MANUSCRIPT 6 four sampling sites, drips 6, 7 and 15 are underlain by stalagmites and lie at the center of the chamber whereas drip 13 lies close to the entrance of the cave and is not associated with a stalagmite. Drip-water was collected in clean plastic bags that were subsequently placed in 1000 ml

T

nalgene bottles. Funnels, 15-cm in diameter, were used to direct the falling drips into the plastic

IP

bags. With the exception below, aliquots of drip-water for 18O and 2H measurements were

CR

stored in 15 ml septum-capped glass serum bottles. The exception occurred during the period

US

December 2012-September 2013 when 8 ml Low Density Polyethylene (LDPE) nalgene bottles were used. In both cases bottles were completely filled to avoid headspace. Aliquots of drip-

AN

water for trace-element measurements were stored in acid-washed 60 ml High Density Polyethylene (HDPE) nalgene bottles. Cave-air temperature was monitored between January and

M

March 2012 and November 2012 to Dec 2013 using an HTAB-176 Certified Hygro/Temp

ED

Indicator and between April and October 2012 using an HOBO Pro v2 Temp/RH Data Logger

PT

(Table 1). Drip flow-rates were estimated by measuring the fluid volume collected during a given time span.

CE

3.2. Laboratory Methods

The rainwater and cave drip-water samples were analyzed on a Gasbench online with a CF-

AC

IRMS (modified Delta-plus). The δ18O and δ2H values of rainwater and drip-water were determined by headspace equilibration methods using CO2 (Epstein and Mayeda, 1953) and H2 (Horita, 1988) pure gases of known isotope compositions, respectively. Final values were normalized to VSMOW/SLAP standards according to Nelson (2000). Isotope values are reported relative to the Vienna Standard Mean Ocean Water (V-SMOW). Analytical precisions for the

ACCEPTED MANUSCRIPT 7 oxygen and hydrogen measurements are ±0.1‰ and ±1.0‰ respectively, based on sample repeats as well as internal and international standards. Drip-water samples were filtered through 0.2-µm pore-size filters and acidified with nitric acid for subsequent major (Ca) and trace-element (Mg and Sr) measurements. Aliquots of several

T

mg by weight, representative of the thin aragonite layers and shaved from an active stalagmite

IP

forming under drip 6, were dissolved in pure nitric acid and the solution was filtered through 0.2

CR

µm pore-size filters. Analyses for the major and trace elements were performed using a Perkin Elmer Optima 3000 DV ICP-OES (Inductively Coupled Plasma – Optical Emission

US

Spectrometer). Calibrations were achieved using multiple element standards from CPI

AN

International Inc. and quality control standards from High Purity Inc. Major and trace-element concentrations are expressed in millimoles/liter (mM). Standard error of the mean for major and

M

trace-element measurements is better than 2% based on standard analysis.

ED

4. Results

4.1. Oxygen and hydrogen isotope compositions of rainfall

PT

The stable isotope composition of rainfall from the rain-gauge station at the University of

CE

Alabama was determined in order to understand the variability in oxygen and hydrogen isotope compositions of cave drip-waters. Table 1 summarizes the monthly averages of rainfall and drip-

AC

water monitoring data (see Supplementary Dataset S1 for a complete listing). Over the 2-year period the rainfall record shows that 2012 was relatively dryer (1405 mm) relative to 2013 (1606 mm) by a factor of 0.87 in annual precipitation (Fig. 2 A). During 2012 monthly rainfall amounts varied from 232 to 16 mm with the highest rainfall in December and the lowest rainfall in April. During 2013 rainfall amounts varied from 256 to 37 mm with the highest rainfall in December

ACCEPTED MANUSCRIPT 8 and the lowest rainfall in May. No significant changes was observed between the monitoring years in both external and in-cave air temperatures (Fig. 2 B). Weekly δ18O and δ2H isotopic compositions (n = 87) varied from 0.8 to -9.5 (‰ V-SMOW) and 8.1 to -66 (‰ V-SMOW) respectively, whereas weighted monthly means δ18O (n = 24) and

T

δ2H (n = 24) varied from -1.2 to -6.4 (‰ V-SMOW) and -4 to -41.6 (‰ V-SMOW),

IP

respectively. Weak interannual 18O and 2H depletion trends correspond with the observed

CR

increase in rainfall amount (Figs. 2 C and 2 D). With one significant exception below, no obvious seasonal trend in rainfall δ18O and δ2H time-series is noticed over the study period. The

US

exception is the weighted monthly mean deuterium excess (d-excess = δ2H - 8δ18O; Craig, 1961)

AN

that exhibits significant seasonality trends with high values during winter (up to 18.5 ‰) and low values during summer (down to 6.2 ‰) (Fig. 2 E).

M

Neither rainfall amount (r2 = 0.012 for δ18O and r2 = 0.004 for δ2H) nor air temperature (r2 =

ED

0.03 for δ18O and r2 = 0.004 for δ2H) seems to control rainwater δ18O and δ2H variability. In contrast, summer low and winter high d-excess values observed in monthly d-excess time-

CE

(Figs. 2 B and 2 E).

PT

series exhibit a significant negative correlation with monthly mean air temperatures (r2 = 0.4)

AC

4.2. Drip-water flow rates and oxygen and hydrogen isotopes The rainfall record from the Childersburg water plant station, which is ~7 km from the DeSoto Caverns, shows that 2012 was relatively dryer than 2013 in terms of total annual precipitation (htpp://www.ncdc.noaa.gov) (Fig. 3 A). The monthly water budget calculated using potential evapotranspiration estimates (Samani, 2000) shows that 2012 had 6 months of moisture-deficit compared to only 2 months in 2013 (Fig 3 B). At the annual scale, the cave dripwater flow rates are substantially higher in 2013 compared to 2012 (Fig. 3 C) and qualitatively

ACCEPTED MANUSCRIPT 9 correlate with the water budget variability. Among the four sampling sites, only drip 6 had continuous discharge throughout the study period with an average flow rate 4.8 times higher in 2013 relative to 2012 (97.3 ml/h in 2012 and 471.3 ml/h in 2013). The slowest drip flow rate (3.0 ml/h) was observed at site 13 in June 2012 whereas the fastest flow rate (3,330 ml/h) was

T

recorded at site 7 in April 2013 (see Supplementary Dataset S2). The higher drip flow rates in

IP

2013 are likely due to the combined effect of increased rainfall amount and decreased potential

CR

evapotranspiration.

Drip-water isotope time series are graphed in Figures 4 A, B, and C. δ18O and δ2H values

US

vary between -5.3 and -3.1 (‰ V-SMOW) and -30.5 and -9.9 (‰ V-SMOW), respectively. The

AN

isotope records from drip 6 show small but significant interannual trends with 18O and 2H relative depletions from the dry/warm 2012 year to the wet/cool 2013 year. The relatively heavy

M

isotope values during 2012 is likely caused by enhance evaporation processes in the epikarst

ED

zone (Bar Matthews et al., 1996). The interannual trends of heavy isotope depletions at the front chamber of the cave are in good agreement with the results of 3-yr (2005-2008) monitoring study

PT

of drip-waters at the back of the cave by Lambert and Aharon, (2010). These authors (op. cit.)

CE

reported interannual trends of 18O and 2H enrichments from a wet year (2005) to a dry year (2008) and attributed the trends to the onset of regional drought conditions. The drip-water d-

AC

excess values lie between 7.9 and 22 ‰ and, unlike the rainfall source, do not exhibit an interannual trend (Fig. 4 C). Unlike the rainfall d-excess values (Fig. 2 C), those of the dripwaters lack a seasonality trend and their mean value of 13.9 ± 0.5 ‰ (n=47) is statistically indistinguishable from the winter rainfall mean value of 14.2 ± 1.6‰ (n=12) suggesting that the cave drip-waters are likely controlled chiefly by winter precipitation.

ACCEPTED MANUSCRIPT 10 The Gulf Coast Meteoric Water Line (GCMWL), established on the basis of monthly weighted mean δ18O and δ2H data (unpublished) from Tuscaloosa precipitation over a decade (2005-2015), is shown in Figure 5. The intercept of the GCMWL equation (8.9 ±1.01) trends towards the global average of 10 (Craig, 1961) but the slope (7.02 ±0.2) is lower than the global

T

average of 8 (Craig, 1961) likely reflecting modifications by secondary processes of re-

IP

evaporation and mixing (Rozanski, et al., 1993). δ18O and δ2H values of rainwater and drip-

CR

waters sampled during the study period fall along the GCMWL line suggesting a common source.

US

4.3. Trace-element concentrations in drip-waters and aragonite speleothems

AN

Trace-element concentrations in drips at DeSoto Caverns vary between 0.30 to 1.59 mM in Ca (mean = 1.19 ± 0.04 mM; n = 61), 1.4 × 10-4 to 5.5 × 10-4 mM in Sr (mean = 4.3 × 10-4 ± 1.1 ×

M

10-5 mM; n = 61) and 0.68 to 1.57 mM in Mg (mean = 1.32 ± 0.03 mM; n = 61) (see

ED

Supplementary Dataset S3). Table 2 lists the mean of trace-elements data from drip site 6 that has a continuous record.

PT

Trace-element compositions of both fresh and weathered bedrock dolomite samples were

CE

analyzed in order to answer the question whether trace-elements in drip-waters are derived from dissolution of pristine or weathered dolomite. Table 3 lists the mean of trace-elements data from

AC

fresh and weathered bedrock dolomite and Figure 6 shows a comparison of drip-waters with weathered and pristine dolomite in terms of Sr/Ca vs Mg/Ca ratios. Here we suggest that the trace elements of cave drip-waters are extracted primarily from dissolution of pristine dolomite based on the proximity of their ratios and significant departure from the ratios of weathered dolomite.

ACCEPTED MANUSCRIPT 11 In general, the time-series of Ca, Sr and Mg concentrations in drip-waters at all four sampling stations (Fig. 1) exhibit an interannual variability coherent with the drip-water δ18O record (Fig. 7). Specifically, the continuous time series at site 6 indicate that delivery of Ca, Mg and Sr to cave drip-water is higher during a “wet” year (2013) than a “dry” year (2012) by a

T

factor of 1.5, 1.3 and 1.4, respectively. Trace-element concentrations are exceptionally low in

IP

June and December 2012 and correspond with the observed increase in δ18O values (Fig. 7). The

CR

observed trends of elemental ratios in drip-waters at DeSoto Caverns corroborate the findings of similar studies in other caves (Fairchild et al., 2000; Tooth and Fairchild, 2003; Cruz et al., 2007;

US

Karmann et al., 2007; Tremaine and Froelich, 2013).

AN

The Mg/Ca and Sr/Ca time-series show interannual trends with slightly higher values during

Mg/Ca ratio in December of 2012.

M

the dry year and lower values during the wet year (Fig. 8 A and 8 B), and an exceptionally high

ED

Elemental concentrations of recently forming aragonite layers from the top of an active stalagmite growing under drip-water 6 are listed in Table 4. Four consecutive aragonite layers

PT

yield mean Mg and Sr concentrations of 37.1 ±11.3 mM and 3.8 ±0.2 mM, respectively, and

CE

Mg/Ca and Sr/Ca molar ratios of 3.84 ±1.16 (x10-3) and 0.392 ±0.09 (x103), respectively. Trace-element distribution coefficients in the aragonite were derived on the basis of

AC

elemental-chemistry data listed in Table 2 and Table 4 of drip-water 6 and the underlain aragonite stalagmite, respectively. ((DTE =(TE/Ca)aragonite/(TE/Ca)solution))

(1)

The distribution coefficients estimated here between drip-water and aragonite for Mg (DMg = 3.49*10-3 ±1.06*10-3) and Sr (DSr = 1.12 ± 0.041) are generally in agreement with the values determined for other karst settings (DMg ≤ 1 and DSr ≥1; Wassenburg et al., 2016).

ACCEPTED MANUSCRIPT 12 5. Discussion 5. 1. Relation between rainfall and drip-water isotopes Alignment of drip-water isotope values with the GCMWL (Fig. 5) offers evidence of their source in the modern-day rainfall. Over the duration of the study, drip-water δ18O and δ2H values

IP

T

(-3.1 to -5.3 ‰ and -9.9 to –30.5 ‰, respectively) exhibit a narrower range by comparison with

CR

the rainfall source (-1.2 to -6.4 ‰ and -4 to -41.6 ‰ respectively, see Figs. 2 & 4). The substantial attenuation of drip-water isotope amplitudes relative to rainfall in the thinly capped,

US

large, front chamber of the cave agrees with the results of a previous 3-year monitoring study from the thickly capped, small, back chamber of the cave (Lambert and Aharon, 2010). The

AN

contrasting isotope compositions between rainfall and drip-water are likely caused by mixing of

M

fresh with residual evaporated water in the epikarst zone (Bar Matthews et al., 1996; Cruz et al.,

ED

2005; Lambert and Aharon, 2010) and bias towards winter precipitation (Aharon and Dhungana, 2017).

PT

The following observations support our contention of an epikarst water storage and seasonal

CE

mixing processes thereof: (i) drip-water 18O and 2H depletion trends (Figs. 4 A and 4 B) are 5% and 11% higher than the trend exhibited by the coeval rainfall (Figs. 2 C and 2 D); (ii) drip-water

AC

mean values of δ18O (-4.6 ± 0.1‰, n =67) and δ2H (-23.0 ± 0.5‰, n =59) are statistically indistinguishable from the weighted means of δ18O (- 4.5 ± 0.4‰, n = 12) and δ2H (- 22.8 ± 2.5‰, n =12) of local winter rainfall suggesting a greater proportion of winter rainfall being stored in the epikarst; (iii) greater groundwater recharge during winter relative to summer months is likely caused by the lesser extent of evapotranspiration in winter relative to summer (Figs. 3 A and 3 B), and (iv) absence of seasonality in drip-water d-excess values stands in sharp contrast to the d-excess of rainfall that exhibit a seasonality with higher values during winter

ACCEPTED MANUSCRIPT 13 relative to summer months. The discrepancy observed between rainfall and drip-water d-excess values are likely explained by the occurrence of post rainfall evaporation in the vadose zone prior to storage in the epikarst (Luo et al., 2012). Lower summer/ higher winter d-excess rainfall values are likely caused by either one, or a

T

combination of the following two factors: (i) maximum (minimum) sub-cloud evaporation

IP

changes related to higher (lower) air temperature (Gat et al., 1996; Lambert and Aharon, 2010;

CR

Guan et al, 2013), and (ii) regional climate system variability in the study area. During summer months, low pressure or trough system builds on the western edge of the Bermuda High causing

US

convective style rainfall while winter rainfall is often associated with cold frontal system that are

AN

driven west to east by jet streams (Lambert and Aharon, 2010). The latter factor is supported by the study of Guan et al. (2013) demonstrating that rainfall d-excess values caused by low

M

pressure or trough systems are significantly lower than that caused by cold frontal systems.

ED

Hence, there is likely a coupling effect of sub-cloud evaporation process and regional climate setting on rainfall d-excess seasonal variations in the study area.

PT

5. 2. Derivation of Mg and Sr Distribution Coefficients in Aragonite Speleothems

CE

Trace-element distribution coefficients between carbonate minerals and solutes have long

AC

been subject of considerable interest because of their dependency on key environmental factors such as ambient temperature, depositional rates, solute chemical composition, among others (McIntyre, 1963; Rimstidt et al., 1998). Proliferation of speleothem studies led to an expanding use of trace-element distribution coefficients for the understanding of paleo-hydroclimate variability (Gascoyne, 1983; Fairchild et al., 2000; Fairchild and Treble, 2009; Tremaine and Froelich, 2013). Previously derived distribution coefficients for calcite DTE(cc) are well understood and have been successfully applied to speleothems whose dominant mineralogy is

ACCEPTED MANUSCRIPT 14 calcite (Huang and Fairchild, 2001; Wassenburg et al., 2016). The application of the DTE(cc) to much rarer aragonite speleothems, however, is impractical because differences in the number of coordination sites between calcite and aragonite (i.e., six-fold vs. nine-fold), dictate differences in cation substitution for Ca, and consequently dissimilar DTE values for the two carbonate

T

polymorphs (Finch and Allison, 2007). Additionally, DTE(ar) values derived from marine biogenic

IP

aragonites (e.g., corals) may not be suitable for freshwater-derived speleothem settings, and

CR

difficulties in precipitating aragonite, maintenance of steady supersaturation under laboratory conditions and dependency on growth rates introduce marked uncertainties in the derivation of

US

reliable DTE(ar) values (Zhong and Mucci, 1989; Dietzel et al., 2004; Gaetani and Cohen, 2006;

AN

Gabitov et al., 2008).

M

There are three ways by which DTE(ar) can be empirically estimated on the basis of aragonite speleothems, using: (i) speleothems consisting of aragonite-calcite coexisting pairs on the

ED

premise that trace elements partitioning into the calcite is well understood (Wassenburg et al.,

PT

2016); (ii) aragonite deposition on synthetic substrates (e.g., glass or ceramics) underlying active drips (Tremaine and Froelich, 2013), and (iii) shaved thin surfaces of actively forming aragonite

CE

stalagmites and the overlying drips, assuming contemporaneity (Finch et al., 2001; 2003).

AC

In this study we used the third method above to estimate the DTE(ar) for Mg and Sr, two of the most commonly used trace elements in speleothem studies, and determined their concentrations (plus common Ca) in four consecutive aragonite layers (Table 4) forming under the perennial drip 6 (Table 2). Our distribution coefficient estimates for Mg and Sr (DMg(ar) = 3.49*10-3±1.06*10-3 and DSr(ar) = 1.12 ± 4*10-2 at ~18 °C) are in good agreement, within the quoted uncertainties, with distribution coefficients for experimentally-derived aragonite (DMg(ar) = 1.7*10-3 ±1.4*10-3 and DSr(ar) = 1.24 ±0.2 at ~ 15 oC (Gaetani and Cohen, 2006). Whereas the Sr

ACCEPTED MANUSCRIPT 15 distribution coefficient reported here agrees well with the empirically-derived Sr estimate from calcite-aragonite pairs in speleothems (DSr(ar) = 1.38 ± 0.53), the distribution coefficient for Mg differs by about two orders of magnitude (DMg(ar) = 9.7*10-5 ±9.01*10-5) (Wassenburg et al., 2016). The general agreement of DMg(ar) ≤ 1 and DSr(ar) ≥ 1 for aragonite stands in sharp contrast

T

to the distribution coefficients in calcite that are some orders of magnitude larger (Mg) and

IP

smaller (Sr) (i.e., DMg(cc) = 1.9*10-2 ±3*10-3 and DSr(cc) = 7.2*10-2 ±1*10-2 at 15 °C; Huang and

CR

Fairchild 2001).

US

5.3 Drip-water trace elements variability

AN

Two principal factors have been proposed to control the variability of trace-element concentrations in drips, and subsequently in speleothems: (i) seasonal change in cave ventilation

M

(Mattey et al., 2010; Wong et al; 2011), and (ii) rainfall amount variability (Fairchild and Treble,

ED

2009; Uchida et al., 2013; Treble et al., 2015). The latter governs change in biomass productivity in the soils overlying caves and/or prior carbonate precipitation in the epikarst zone (Fairchild et

PT

al., 2000; Sinclair, 2011). Factors controlling the trace elements variability in the drip-water

CE

reported in this study are assessed below. Although cave pCO2 data at DeSoto Caverns exhibit pronounced seasonal changes in

AC

ventilation (Dhungana, 2017), absence of seasonal variability evidence in the Ca, Sr and Mg time-series (Fig. 7) argues against a seasonal change in cave ventilation as a principal controlling factor. Instead, we contend that rainfall-amount variability exerts a dominant control on the elemental concentrations and their ratios in the drips through the coupling of biomass productivity overlying the cave and PAP variability in the epikarst. The following lines of evidence corroborate our contention. First, the higher elemental concentrations in the “wet” year (2013) relative to the “dry” year (2012) are likely linked to the relatively higher biomass

ACCEPTED MANUSCRIPT 16 productivity caused by higher precipitation (Fig. 7). The enhanced plant-root respiration during higher vegetation growth increases soil-CO2 concentrations and its uptake in the fluid moving through the epikarst causing enhanced dissolution of carbonates (Li et al., 2011). Second, PAP formed up-stream from the drips likely affects the trace elemental concentrations and

T

consequently the Mg/Ca and Sr/Ca ratios in the drip-water emerging on the cave ceiling, and

IP

thus the speleothems. This is because distribution coefficients of trace-element ratios during

CR

dissolution show strong preference for the fluid phase (Fairchild et al., 2000; Treble et al., 2015). Significant removal of Ca relative to Mg and Sr from seepage water during the PAP process at

US

DeSoto Caverns will increase both Mg/Ca and Sr/Ca ratios in the cave drip-waters.

AN

PAP likely occurs mostly during drought events because of aragonite deposition in cavities with a lower pCO2 than the seeping water (Tooth and Fairchild, 2003; Johnson et al., 2006;

M

McDonald et al., 2007). Thus, relatively low Ca, Mg and Sr concentrations (Fig. 7) and high

ED

Mg/Ca and Sr/Ca ratio time-series during 2012 relative to 2013 likely document the transition from a “dry” to a “wet” year (Fig. 8). Specifically, during the “dry” year the elemental

PT

concentrations were lower by about factor of 1.5 and the trace-element ratios were higher by

CE

about factor of 1.2 relative to the “wet” year (Figs. 7 and 8). Finally, the observed positive linear relationship between Sr and Mg concentrations (Fig. 9) offers additional evidence of PAP in the

AC

epikarst causing Ca removal and coeval Sr and Mg enrichments in the drip-waters.

ACCEPTED MANUSCRIPT 17 6. Conclusions Figure 10 summarizes the results of this investigation aiming to provide insights on hydroclimate proxies contrast between years with distinct rainfall amounts variability. The principal conclusions of the study are as follows.

T

1. Over the two years study of the DeSoto Caverns, cave air temperature was stable and relative

IP

humidity was >95 %. In contrast, rainfall amounts were highly variable between the years. Drip-

CR

water δ18O and δ2H values increased during the dry year and decreased during the wet year while elemental concentrations decreased (increased) and their ratios increased (decreased) during the

US

dry (wet) year.

AN

2. Observed interannual variability in trace-element concentrations and Mg/Ca and Sr/Ca ratios are likely controlled by the rainfall amount variability. This finding corroborates the results of

M

Tremaine and Froelich (2013) at Hollow Ridge Cave who reported the dependency of

ED

interannual variations in Mg/ Ca and Sr/ Ca ratios on rainfall amount variability. 3. Drip-waters stable isotope and trace elemental data reported here have important implications

PT

for the speleothem-based paleo-hydroclimate interpretations because the hydroclimate proxies

CE

are likely to be archived in the speleothems deposited underneath the drips (Cruz et al., 2007; Mattey et al., 2010; Wong et al., 2011).

AC

7. Acknowledgments

We thank Tim Lacy and Allen Mathis III for providing access to the cave for observations and samples collection, and graduate and undergraduate students who assisted in fieldwork. J. W. Lambert is thanked for advice and assistance in the Alabama Stable Isotope Laboratory (ASIL), and S. Bhattacharyya for assistance with elemental chemistry measurements. We thank the two journal reviewers, anonymous and Dr. Bogdan Onac, for their constructive criticisms.

ACCEPTED MANUSCRIPT 18 This study was funded by the Gulf Coast Association of Geological Societies (GCAGS) faculty and student grants, and multiple University of Alabama student research grants to the senior author. 8. References

IP

T

Aharon, P., Dhungana, R., 2017. Ocean-atmosphere interactions as drivers of mid-to-late

Caverns, Southeast USA. Quat. Sci. Rev. 170, 69-81.

CR

Holocene rapid climate changes: Evidence from high-resolution stalagmite records at DeSoto

US

Baigorria, G.A., Jones, J.W., Shin, D.W., Mishra, A., O’Brien, J.J., 2007: Assessing

Model outputs. Clim. Res. 34, 211-222.

AN

uncertainties in crop model simulations using daily bias-corrected Regional Circulation

M

Bar-Matthews M., Ayalon A., Matthews A., Sass E., Halicz L., 1996. Carbon and oxygen isotope study of the active water-carbonate system in a karstic Mediterranean cave:

ED

implications for paleoclimate research in semiarid regions. Geochim. Cosmochim. Acta 60,

PT

337–347.

Burns, S.J., Flietmann, D., Matter, A., Neff, U., Mangini, A., 2001. Speleothem evidence from

CE

Oman for continental pluvial events during interglacial periods. Geology. 29, 623-626.

AC

Burns, S.J., Fleitmann, D., Mudelsee, M., Neff, U., Matter, A., Mangini, A., 2002. A 780-year annually resolved record of Indian Ocean monsoon precipitation from a speleothem from south Oman. J. Geophys. Res-Atm. 107 (D20), 4434. Craig, H., 1961. Isotopic Variations in Meteoric Waters. Science. 133 (3465), 1702-1703. Cruz, F.W., Karmann, I., Viana, O., Burns, S. J., Ferrari, J. A., Vuille, M., Sial, A.N., Moreira, M. Z., 2005. Stable isotope study of cave percolation waters in subtropical Brazil: Implications for paleoclimate interferences from speleothems, Chem. Geol., 220, 245-262.

ACCEPTED MANUSCRIPT 19 Cruz, F.W., Burns, S. J., Jercinovic, M., Karmann, I., Sharp, W.D., Vuille, M., 2007. Evidence of rainfall variations in Southern Brazil from trace-element ratios (Mg/Ca and Sr/Ca) in a Late Pleistocene stalagmite. Geochim. Cosmochim. Acta 71 (9), 2250-2263. Dietzel, M., Gussone, N., Eisenhauer, A., 2004. Co-precipitation of Sr2+ and Ba2+ with aragonite

T

by membrane diffusion of CO2 between 10 and 50 °C. Chem. Geol. 203 (1-2), 139-151.

IP

Dhungana, R., 2017. Cave-air CO2 and drip-water geochemical variability at DeSoto Caverns:

CR

Implications for speleothem-based paleoclimate studies. PhD dissertation, University of Alabama, pp. 219.

US

Epstein, S., Mayeda, T., 1953. Variation of 18O content of waters from natural sources. Geochim.

AN

Cosmochim. Acta, 5 (5), 213-224.

Fairchild, I.J., Borsato, A., Tooth, A.F., Frisia, S., Hawkesworth, C.J., Huang, Y., McDermott,

M

F., Spiro, B., 2000. Controls on trace element (Sr-Mg) compositions of carbonate cave

ED

waters: implications for speleothem climatic record. Chem. Geol. 166 (3-4), 255-269. Fairchild, I.J., Treble, P.C., 2009. Trace elements in speleothems as recorders of environmental

PT

change. Quat. Sci. Rev. 28 (5-6), 449-468.

CE

Feng, S., Hu, Q., Wu, Q., Mann, M.E., 2013. A Gridded Reconstruction of Warm Season Precipitation for Asia Spanning the Past Half Millennium. J. Clim. 26, 2192-2204.

AC

Finch, A. A., Shaw, P. A., Weedon, G. P., Holmgren, K., 2001. Trace-element variation in speleothem aragonite: potential for palaeoenvironmental reconstruction. Earth Planet. Sci. Lett. 186, 255-267. Finch, A. A., Shaw, P. A., Holmgren, K., Lee-Thorp, J., 2003. Corroborated rainfall records from aragonitic stalagmites. Earth Planet. Sci. Lett. 215, 265-273.

ACCEPTED MANUSCRIPT 20 Finch, A. A., Allison, N., 2007. Coordination of Sr and Mg in calcite and aragonite. Mineral. Mag. 71 (5), 539-552. Fleitmann, D., Burns, S.J., Mudelsee, M., Neff, U., Kramers, H., Mangini, A., Matter, A., 2003. Holocene Forcing of the Indian Monsoon Recorded in a Stalagmite from Southern Oman.

T

Science 300, 1737-1739.

IP

Gabitov, R. I., Gaetani, G. A., Watson, E. B., Cohen, A. L., Ehrlich, H. L., 2008. Experimental

CR

determination of growth rate effect on U6+ and Mg2+ distribution between aragonite and fluid at elevated U6+ concentration. Geochim. Cosmochim. Acta 72, 4058-4068.

US

Gascoyne, M., 1983. Trace-element partition coefficients in the calcite-water system and their

AN

paleoclimatic significance in cave studies. J. Hydrol. 61, 213-222. Gat, J.R., Shemesh, A., Tziperman, Eli., Hecht, A., Georgopoulos, D., Basturk, O., 1996. The

M

stable isotope composition of waters of the eastern Mediterranean Sea. J. Geophys. Res. 101

ED

(C3), 6441-6451.

Gaetani, G.A., Cohen, A.L., 2006. Element partitioning during precipitation of aragonite from

PT

seawater: A framework for understanding paleoproxies. Geochim. Cosmochim. Acta 70 (18),

CE

4617-4634.

Guan, H., Zhang, X., Skrzypek, G., Sun, Z., Xu, X., 2013. Deuterium excess variations of

AC

rainfall events in a coastal area of South Australia and its relationship with synoptic weather systems and atmospheric moisture sources. J. Geophys. Res. 118, 1123-1138. Horita, J., 1988. Hydrogen isotope analysis of natural waters using an H2 – water equilibrium method: a special application to brines. Chem. Geol (Iso. Geosc. Sec.) 72, 89-94. htpp://www.ncdc.noaa.gov. Accessed on March 14, 2017.

ACCEPTED MANUSCRIPT 21 Huang, Y., Fairchild, I.J., 2001. Partitioning of Sr2+ and Mg2+ into calcite under karst-analogue experimental conditions. Geochim. Cosmochim. Acta 65 (1), 47-62. Johnson, K.R., Hu, C., Belshaw, N.S., Henderson, G.M., 2006. Seasonal trace-element and stable-isotope variations in a Chinese speleothem: The potential for high-resolution

T

paleomonsoon reconstruction. Earth Planet. Sci. Lett. 244 (1-2), 394-407.

IP

Karmann, I., Cruz, F.W., Viana, O., Burns, S.J., 2007. Climate influence on geochemistry

CR

parameters of waters from Santana-Pérolas cave system, Brazil. Chem. Geol. 244 (1-2), 232247.

AN

values. Quat. Sci. Rev. 28 (5-6), 412-432.

US

Lachniet. M.S., 2009. Climatic and environmental controls on speleothem oxygen-isotope

Lambert, W.J., Aharon, P., 2010. Oxygen and hydrogen isotopes of rainfall and drip water at

M

DeSoto Caverns (Alabama, USA): Key to understanding past variability of moisture

ED

transport from the Gulf of Mexico. Geochim. Cosmochim. Acta 74, 846-861. Li, T-Y., Shen, C-C., Li, H-C., Li, J-Y., Chiang, H-W., Song, S-R., Yuan, D-X., Lin, C. D-J.,

PT

Gao, P., Zhou, L., Wang, J-L., Ye, M-Y., Tang, L-L., Xie, S-Y., 2011. Oxygen and carbon

CE

isotope systematics of aragonite speleothems and water in Furing cave, Chongqing, China. Geochim. Cosmochim. Acta 75 (15), 4140-4156.

AC

Luo, W., Wang, S., Xie, X., 2012. A comparative study on the stable isotopes from precipitation to speleothems in four caves of Guizhou, China. Chem. der Erde- Geochem. 73 (2), 205-215. Mattey, D., Lowry, D., Duffet, J., Fisher, R., Hodge, E., Frisia, S., 2008. A 53 year seasonally resolved oxygen and carbon isotope record from a modern Gibraltar speleothem: Reconstructed drip water and relationship to local precipitation. Earth Planet. Sci. Lett. 269, 80-95.

ACCEPTED MANUSCRIPT 22 Mattey, D.P., Fairchild, I.J., Atkinson, T.C., Latin, J.P., Ainsworth, M., Durell, R., 2010. Seasonal microclimate control of calcite fabrics, stable isotopes and trace elements in modern speleothem from Michaels Cave, Gibraltar. Geol. Soc. Spec. Pub. (London) 336, 323-344. McDonald, J., Drysdale, R., Hill, D., Chisari, R., Wong, H., 2007. The hydrochemical response

T

of cave drip waters to sub-annual and inter-annual climate variability, Wombeyan Caves, SE

IP

Australia. Chem. Geol. 244 (3-4), 605-623.

CR

McIntire, W. L., 1963. Trace element partition coefficients-A review of theory and application to geology. Geochim. Cosmochim. Acta 27 (12), 1209-1264.

US

Nelson, S.T., 2000. A simple, practical methodology for routine VSMOW/SLAP normalization

AN

of water samples analyzed by continuous flow methods. Rapid Commun. Mass Spec. 14, 1044-1046.

M

Ortegren, J.T., Knapp, P.A., Maxwell, J.T., Tyminski, W.P., Soulé, P.T., 2011. Ocean-

ED

atmosphere influences on low-frequency warm-season drought variability in the Gulf Coast and Southeastern Unites States. J. App. Meteorol. Clim. 1177-1186, doi:

PT

10.1175/2010JAMC2566.1.

CE

Oster, J.L., Montañez, I.P., and Kelley, N.P., 2012. Response of a modern cave system to large seasonal precipitation variability. Geochim. Cosmochim. Acta 91, 92-108.

AC

Riechelmann, S., Schröder-Ritzrau, A., Wassenburg, J. A., Schreuer, J., Richter, D. K., Riechelmann, D. F., Terente, M., Constantin, S., Mangini, A., Immenhauser, A., 2014. Physicochemical characteristics of dripwaters: influence on mineralogy and crystal morphology of recent cave deposits precipitates. Geochim. Cosmochim. Acta 145, 13-29. Rozanski, K., Araguás-Araguás, L., Gonfiantini, R., 1993. Isotopic Patterns in Modern Global Precipitation, in: Swart, P.K., Lohmann, K.C., McKenzie, J., Savin, S., (Eds.), Climate

ACCEPTED MANUSCRIPT 23 Change in Continental Isotopic Records. Geophys. Mon. Ser. 78, Am. Geophys. Union., Washington, DC., pp. 1-36. Rimstidt, J. D., Balog, A., Wobb, J., 1998. Distribution of trace elements between carbonate minerals and aqueous solutions. Geochim. Cosmochim. Acta 62 (11), 1851-1863.

T

Samani, Z., 2000. Estimating solar radiation and evapotranspiration using minimum

IP

climatological data. J. Irrig. Drain. Eng. 126, 265-267.

CR

Sinclair, D.J., 2011. Two mathematical models of Mg and Sr partitioning into solution during incongruent calcite dissolution. Implications for drip water and speleothem studies. Chem,

US

Geol. 283, 119-133.

AN

Tooth, A.F., Fairchild, I.J., 2003. Soil and karst aquifer hydrological controls on the geochemical evolution of speleothem-forming drip waters, Crag Cave, southwest Ireland. J. Hydrol. 273

M

(1-4), 51-68.

ED

Treble, P.C., Fairchild, I.J., Griffiths, A., Baker, A., Meredith, K.T., Wood, A., McGuire, E., 2015. Impacts of cave air ventilation and in-cave prior calcite precipitation on Golgotha Cave

PT

drip water chemistry, southwest Australia. Quat. Sci. Rev.127, 61-72.

CE

Tremaine, D.M., Froelich, P.N. and Wang, Y., 2011. Speleothem calcite farmed in situ: Modern calibration of δ18O and δ13C paleoclimate proxies in a continuously-monitored natural cave

AC

system. Geochim. Cosmochim. Acta 75 (17), 4929-4950. Tremaine, D.M., Froelich, P.N., 2013. Speleothem trace element signatures: A hydrologic geochemical study of modern cave drip waters and farmed calcite. Geochim. Cosmochim. Acta 121, 522-545. Uchida, S., Kurisaki, K., Ishiro, Y., Haraguchi, S., Yamanaka, T., Noto, M., Yoshimura, K., 2013. Anthropogenic impact records of nature for past hundred years extracted from

ACCEPTED MANUSCRIPT 24 stalagmites in caves found in the Nanatsugama sandstone formation, Saikai, Southwestern Japan. Chem. Geol. 347, 59-68. Wang, Y., Cheng, H., Edwards, L.R., Kong, X., Shao, X., Chen, S., Wu, J., Jiang, X., Wang, X., An, Z., 2008. Millennial- and orbital-scale changes in the East Asian monsoon over the past

T

224,000 years. Nature 451, 1090-1093.

IP

Wassenburg, J. A., Scholz, D., Jochum, K. P., Cheng, H., Oster, J., Immenhauser, A., Richter, D.

CR

K., Häger, T., Jamieson, R. A., Baldini, J. U. L., Hoffmann, D., Breitenbach, S. F. M., 2016. Determination of aragonite trace element distribution coefficients from speleothem calcite-

US

aragonite transitions. Geochim. Cosmochim. Acta 190, 347-367.

AN

Wong, C.I., Banner, J.L., Musgrove, M., 2011. Seasonal drip water Mg/Ca and Sr/Ca variations driven by cave ventilation: Implications for and modeling of speleothem paleoclimate

M

records. Geochim. Cosmochim. Acta 75 (12), 3514-3529.

ED

Zhong, S., Mucci, A., 1989. Calcite and aragonite precipitation from seawater solutions of various salinities: Precipitation rates and overgrowth compositions. Chem. Geol. 78, 283-

AC

CE

PT

299.

ACCEPTED MANUSCRIPT 25 Figure Captions Fig. 1. Plan view of DeSoto Caverns (modified from Lambert and Aharon, 2010) showing the four drip-water sampling locations: site 6 (solid black circle), site 7 (solid green square), site 13 (solid blue square) and site 15 (solid red circle). The inset shows the map of southeastern US

IP

T

with the locations of DeSoto Caverns (solid green star) and the rain-gauge station on the campus

CR

of the University of Alabama in Tuscaloosa, Alabama (solid yellow circle).

Fig. 2. Time series of monitoring data during 2012-2013 years. Uncertainty bars are standard

US

error of the mean. Monthly precipitation from the rain-gauge at the University of Alabama in Tuscaloosa. (B) Monthly mean air temperature recorded at Childersburg water plant station

AN

(www.ncdc.noaa.gov) and superimposed cave air temperature. (C) Weighted monthly δ18O mean

M

of rainwater. (D) Weighted monthly δ2H mean of rainwater. (E) Weighted monthly d-excess

ED

mean of rainwater. Modest interannual 18O and 2H depletion trends, marked by an interrupted line, are observed in δ18O and δ2H time series. Complete data are listed in Supplementary

PT

Dataset S1.

CE

Fig. 3. Drip-flow rates are compared to coeval rainfall parameters. (A) Rainfall amount record at Childersburg water plant station from January 2012 to December 2013 (www.ncdc.noaa.gov).

AC

(B) Net water budget after accounting for evapotranspiration (ET) using an equation from Samani (2000). (C) Drip-flow rates (ml/h) for all four monitored drips. Refer to Fig. 1 for drip site locations and symbols. Complete data are listed in Supplementary Dataset S2. Fig. 4. Time series of isotope composition of drip-water at DeSoto Caverns. Interrupted lines in (A) and (B) mark interannual negative δ18O and δ2H trends for site 6. Refer to Fig. 1 for drip site

ACCEPTED MANUSCRIPT 26 locations and symbols. (A) δ18O time series; (B) δ2H time series, and (C) d-excess time series. Complete data are listed in Supplementary Dataset S2. Fig. 5. Cross plot of oxygen and hydrogen isotope compositions of rainfall and drips measured in this study. Open circles and triangles are the isotope compositions of rainwater during 2012 and

IP

T

2013, respectively. Refer to Fig. 1 for drip site locations and symbols. . Both rain and drip samples line-up along the Gulf Coast Meteoric Water Line (GCMWL) that is established on the

CR

basis on monthly weighted mean δ18O and δ2H decadal data (2005-2015, unpublished) at the

US

University of Alabama gauge station. The GCMWL regression equation is: δ2H = (8.9 ±1.01) + (7.02 ±0.2) δ18O (r2 = 0.91).

AN

Fig. 6. Cross plot of Sr/Ca and Mg/Ca ratios of drips and weathered (WD) and pristine (PD)

M

dolomites at DeSoto Caverns. The proximity of the trace-element ratios of drip-water to that of

ED

pristine dolomite field suggests that dissolution of the latter forms the source of trace elements in the drips. Complete data of the drip-waters are listed in Supplementary Dataset S3.

PT

Fig. 7. In-phase variations of trace elements and δ18O time series in drip-water at DeSoto

CE

Caverns. (A) Ca time series; (B) Sr time series; (C) Mg time series, and (D) δ18O time series. Symbols representing drip-water sites as in Figure 1. Time series data for drip-water at site 6 are

AC

connected by lines.

Fig. 8. Time series of drip-water trace-element ratios at DeSoto Caverns. Symbols representing drip-water sites as in Figure 1. Time series data for drip-water at site 6 are connected by lines. (A) Mg/Ca ratios; (B) Sr/Ca ratios. Time series of Mg/Ca and Sr/Ca exhibit similar interannual phase changes, with higher ratios during the 2012 dry year and lower ratios during the 2013 wet year.

ACCEPTED MANUSCRIPT 27 Fig. 9. Cross plots of Sr vs. Mg concentrations in DeSoto Caverns drip-water. Symbols representing drip-water sites as in Figure 1. Regression lines of individual drip sites exhibit statistically significant positive correlations between Sr and Mg, suggesting prior aragonite precipitation (PAP) upstream of the drips that are feeding the stalagmites.

IP

T

Fig. 10. Cartoon of the karst landscape at DeSoto Caverns summarizing the factors controlling the geochemistry of cave drip-water and speleothems. (A) Factors controlling the geochemistry

CR

of drip- water during dry/warm conditions, and (B) Factors controlling the geochemistry of drip-

AC

CE

PT

ED

M

AN

US

water during wet/cool conditions.

ACCEPTED MANUSCRIPT 28

Table 1. Averages of rainfall and drip water monitoring data. Uncertainties are standard error of the mean (1σ). Drip water data are from drip site 6 (Fig. 1), which has a continuous flow record and is representative of the other front chamber drips. Data other than Drip 6 are listed in Appendix 2. Outside

cave T

δ18Orain

δ2Hrain

fall

rate

cave T

(°C)

(‰ V-

(‰ V-

(mm)

(ml/h)

(°C)

2-Year

1505

301

17.8

18.2

-4.4 ±0.3

(2012-

±29

±57.4

±1.4

±0.1

(n=24)

(n=21)

(n=24)

(n=23)

97 ±26

18.8

17.9

(n=9)

±1.9

2012

±0.1

CR SMOW)

US

-23.6 ±2.0

δ2Hdrips

(‰ V-

(‰ V-

SMOW)

SMOW)

-4.7 ±0.1

-24.3 ±0.6

(n=23)

(n=22)

(n=24)

-4.4 ±0.4

-23.0

-4.5 ±0.1

-22.4 ±0.7

(n=12)

±2.8

(n=11)

(n=10)

(n=12)

(n=11)

16.8

18.5

-4.5 ±0.3

-24.2

-4.8 ±0.1

-25.8 ±0.6

(n=12)

±2.9

(n=12)

(n=12)

PT

471 ±69.3

±2.0

±0.2

(n=12)

(n=12)

(n=12)

CE

2013

1606

AC

Year

AN

1405

M

Year

ED

2013)

SMOW)

δ18Odrips

T

Drip

IP

Rain

Period

(n=12)

(n=12)

ACCEPTED MANUSCRIPT 29 Table 2. Drip-water trace-element concentrations from site 6. Uncertainties are standard error of the mean (1σ). Ca2+

Mg2+

Sr2+

(mM)

(mM)

(mM ×104)

2-Year (2012-

1.2 ±0.1

1.3 ±0.04

4.2 ±0.2

1.1 ±0.04

0.35 ±0.01

2013)

(n=22)

(n=22)

(n=22)

(n=22)

(n=22)

Year

1.0 ±0.1

1.1 ±0.03

3.5 ±0.2

1.2 ±0.1

0.37 ±0.01

2012

(n=11)

(n=11)

(n=11)

(n=11)

(n=11)

1.4 ±0.03

1.5 ±0.03

4.8 ±0.1

1.0 ±0.03

0.33 ±0.01

(n=11)

(n=11)

(n=11)

(n=11)

(n=11)

CE

PT

ED

T

IP

CR

M

AN

2013

AC

Year

Mg/Ca

US

Period

Sr/Ca (× 103)

ACCEPTED MANUSCRIPT 30

Table 3. Trace-element data from pristine and weathered bedrock dolomite. Uncertainties are standard error of the mean (1σ).

(n=3)

(n=3)

M ED

T 4895.9 ±46.9

1.1 ±0.02 (n=3)

Sr2+/Ca

0.98 ±0.003

0.38 ±0.001

(n=3)

(n=3)

0.96 ±0.01

0.21 ±0.004

(n=3)

(n=3)

IP

5084.7 ±97.1

Mg/Ca

(× 103)

CR

(n=3)

1.7 ±0.004 (n=3)

US

(n=3)

Sr2+ (mM)

AN

4334.1 ±14.9

PT

Weathered

Mg2+ (mM)

4416.8 ±25.6

CE

Pristine

Ca2+ (mM)

AC

Dolomite

ACCEPTED MANUSCRIPT 31

Table 4. Trace-element concentrations of recently formed aragonites from top of an actively stalagmite growing underneath drip 6. Uncertainties are standard error of the mean (1σ). Mg2+ (mM)

Sr2+ (mM)

0.0046

0.3732

0.0018

0.4001

3.7

0.0068

0.3816

4.0

0.0023

0.4130

3.8 ±0.2

0.00384

0.3920

±0.00116

±0.0899

44

3.6

D6B

9496.8

16.7

3.8

D6C

9694.8

65.8

D6D

9684.5

21.8

9631 ±46

37.1 ±11.3

US AN

M

ED PT CE

CR

9646.3

AC

Sr/Ca (× 103)

D6A

Mean

Mg/Ca

T

Ca2+ (mM)

IP

Sample

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10