This Article Appeared in a Journal Published by Elsevier. the Attached

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

Available online at www.sciencedirect.com

Geochimica et Cosmochimica Acta 75 (2011) 6859–6869 www.elsevier.com/locate/gca

Paleo-environmental implication of clumped isotopes in land snail shells

Shikma Zaarur, Gerard Olack 1, Hagit P. Aﬀek ⇑

Department of Geology and Geophysics, Yale University, CT, USA

Received 10 December 2010; accepted in revised form 31 August 2011

Abstract

Clumped isotopes analyses in modern land snail shells are reported and used to interpret shell oxygen isotopes within the context of terrestrial paleo-climatology. Carbonate clumped isotopes thermometry is a new technique for estimating the temperature of formation of carbonate minerals. It is most powerful as an indicator of environmental parameters in combination with d18O, allowing the partitioning of the d18O signal into its temperature and water components. Results indicate that snail shell calcification temperatures are typically higher than either the mean annual or the snail activity season ambient temperatures. Small inter- and intra-snail variability suggests that shell aragonite forms at isotopic equilibrium so that the derived temperatures are an eco-physiological parameter reflecting snail body temperature at the time of calcification. We attribute these higher body temperatures to snail eco-physiological adaptations through shell color, morphology, and behavior. In combination with shell oxygen isotope composition, these temperatures allow us to calculate snail body water composition, which is in turn interpreted as a paleo-hydrological indicator, reflecting isotopic composition of local precipitation modified by local evaporation. Ó 2011 Elsevier Ltd. All rights reserved.

1. INTRODUCTION and temperature changes (Goodfriend, 1992). Oxygen isotope composition of carbonate minerals is the most popular Land snails have been used as an archive material for proxy in paleo-climate research, used in both marine and reconstructing past climatic and environmental conditions terrestrial environments. It reflects the temperature depen- on land through eco-physiological analysis of snail species dent equilibrium isotopic fractionation between the mineral growth habitats, as well as through isotopic analysis of snail and the water in which it grows (e.g., Epstein et al., 1953). shell aragonite (Goodfriend, 1992). The occurrence of However, implementation of d18O for reconstruction of extant species in the Quaternary fossil record allows the environmental temperatures requires an independent esti- use of modern analog approaches to estimate paleo-envi- mate of paleo-water composition, which is especially diffi- ronmental conditions. In particular, the isotopic composi- cult in terrestrial environments, due to the complexity and tion of snail shells together with snail faunal assemblages variability of rain-water isotopic composition. and shell morphology have been used to reconstruct The aragonitic snail shell precipitates from bicarbonate environmental changes such as vegetation type coverage dissolved in the snail body fluids, and is assumed to be in isotopic equilibrium with body water (Goodfriend, 1992). ⇑ The main source for land snail body water is rain and Corresponding author. Address: Department of Geology and dew that penetrate through the integument of the foot Geophysics, Yale University, 210 Whitney Ave., P.O. Box 208109, (Prior, 1985; Heller, 1993). d18O is therefore predominantly New Haven, CT 06520-8109, USA. Tel.: +1 203 432 3761. E-mail addresses: [email protected] (S. Zaarur), golack controlled by isotopic composition of the local rain (Bala- @uchicago.edu (G. Olack), hagit.aff[email protected] (H.P. Affek). krishnan and Yapp, 2004) and dew (Goodfriend et al., 1 Present address: Department of Geophysical Sciences, Univer- 1989) with potential effects of evaporation at the soil sity of Chicago, IL, USA. surface. The life span of land snails ranges from several

6860 S. Zaarur et al. / Geochimica et Cosmochimica Acta 75 (2011) 6859–6869 months to 19 years (Barker, 2001) so that shell d18O may be of 13C and 18O in the carbonate mineral. When the mineral associated with a specific season in small snails but in larger precipitates under conditions of isotopic equilibrium this shells is likely to reflect precipitation over a number of preference is temperature dependent and therefore records years. the temperature of mineral growth, independent of the oxy- The first study of stable isotopes in land snail shells gen isotope composition of the solution in which it grows (Yapp, 1979) was followed by many more (Magaritz and (Wang et al., 2004; Ghosh et al., 2006; Eiler, 2007). The Heller, 1980; Lecolle, 1985; Goodfriend et al., 1989; D47–T relationship was first calibrated by Ghosh et al. Goodfriend, 1990, 1992; Leng et al., 1998; Leone et al., (2006) through analysis of synthetic calcite and was then ex- 2000; Goodfriend and Ellis, 2002; Balakrishnan and Yapp, panded to biogenic materials (summarized by Eiler (2007) 2004; Balakrishnan et al., 2005; Zanchetta et al., 2005; Li and Tripati et al. (2010)). The temperatures derived from et al., 2007; Yanes et al., 2008, 2009; Kehrwald et al., clumped isotopes can then be combined with carbonate 2010) which investigated correlations between carbon and d18O, to obtain the oxygen isotopic compositions of the oxygen isotope compositions and environmental condi- water in which the mineral formed. tions. The approach was based on extrapolation from stud- Published data on aragonitic material of biogenic origin, ies in aquatic snail shell carbonates whose d18O values, both argonitic and calcitic, is in general agreement with the similarly to many marine biogenic carbonates, reflect the calcite calibration (Schauble et al., 2006; Eiler, 2007; Tripati temperatures and isotopic composition of lake waters, et al., 2010). However, using the same approach in land and assuming a similar relationship in non-aquatic land snails we observed deviations from D47 values expected snails. The pioneering work of Yapp (1979) and Magaritz from ambient environmental temperatures, suggesting that and Heller (1980) examined d18O of land snail shells as a snail calcification temperatures are higher than those in the proxy for climatic and environmental conditions (mainly snail growth habitat, with the temperature offset being lar- rainfall d18O). Even though no simple correlation was ger at lower environmental temperatures. The calcification found between the snail shells and the estimated meteoric temperatures obtained are nevertheless useful for ecological water d18O, they identified relative humidity as a key and physiological studies of snails, and when considered to- parameter in relating rainfall to snail body water gether with shell d18O, for determining snail body water iso- composition. topic compositions. In later studies, for example in France and Italy (Lecolle, 1985; Zanchetta et al., 2005), good correlations 2. MATERIALS AND METHODS between shell d18O values and that of the environmental water were observed. These correlations did not hold, how- Modern snail shells were collected in a variety of loca- ever, at high altitudes (1200 m above sea level) and in re- tions worldwide, covering a large range of environmental gions where the mean annual temperature was lower than conditions, from hot and dry (e.g., Negev, Israel) to cold 6 °C(Lecolle, 1985). Although several studies established and humid (e.g., Davos, Switzerland) (Table 1). Specimens a correlation between d18O of snail shell carbonate and lo- were either collected by us or obtained from the Yale Pea- cal precipitation, the failure of these relationships at low body Museum invertebrate collection. Additional sample temperature regions (Lecolle, 1985) remains a significant information can be found in Supplementary Information. source of uncertainty in extrapolating snail shell d18O data The samples were broken or empty shells collected in the in both space and time. vicinity of live active snails. Samples include a variety of d13C signatures of shell carbonates and organic matter species, with the same species available in several locations reflect the carbon sources of snail diet (Goodfriend, (Helicidae Helix aspersa from Palo Alto and San Simeon in 1992), being indicative of vegetation types in the immediate California, and Haifa in the north of Israel). Different spe- snail environment (Goodfriend and Magaritz, 1987; Good- cies growing in one location were compared as well (Pleu- friend and Ellis, 2002; Stott, 2002; Metref et al., 2003; Bal- rodonte acuta and Orthalicus undatus from Jamaica and akrishnan and Yapp, 2004). Carbon isotope records are Sphincterochila zonata and Trochoidea simulate from the thus interpreted to reflect variations in C3 and C4 plant dis- Negev, in the south of Israel). tribution in response to changes in rainfall patterns (Good- In preparation for isotopic analysis, shells were sliced friend and Magaritz, 1987; Goodfriend, 1990). A vertically using a Buehler Isomet Low Speed Bone Saw, complication, however, is potential carbon contribution to allow removal of remaining organic tissue and soil par- to the shell from ingestion of carbonate rocks and atmo- ticles. Shells were rinsed in DI water, and cleaned by son- spheric CO2 (Goodfriend, 1992). ication (Sonicor DSC-100TH). Shell material was dried in We examine clumped isotopes in modern land snail shells vacuum at room temperature and then ground in an agate and use them to test the assumptions governing d18O mortar and pestle. All samples were identified as arago- interpretations. The carbonate clumped isotopes paleo- nitic by X-ray diffraction (using an automated Scintag thermometry technique is based on the abundance of PAD V diffractometer, in the XRD laboratory at Yale 13C–18O bonds in the carbonate lattice relative to that ex- University). pected at a random distribution of isotopes among all isoto- Shell carbonate was digested overnight in 105% H3PO4 3 pologues (quantified by the parameter D47; Eiler, 2007). As (q = 1.93 g/cm )at25°C. The CO2 extracted was analyzed 18 13 such it measures the thermodynamically controlled for d O, d C and D47.CO2 was cryogenically extracted on preference of two heavy isotopes to create a bond with each a vacuum line and then cleaned by passing through a GC other, and it does not depend on the absolute concentrations column (Supelco Q-Plot, 30 m long, 0.53 mm ID) at Author's personal copy

Table 1a 18 18 Compiled data for all samples. d OSBW is the snail body water calculated using either D47 or the environmental temperature. d Op is the isotopic composition of local precipitation (Bowen and Revenaugh, 2003). All d18O values of water are referenced to VSMOW. Shell color codes are as follows: D = dark, M = Medium, B = brown and W = white. 13 18 18 18 18 Sample Location n Species d C(&) d O(&) D47 (&) D47 (&) d OSBW d OSBW (env. d OP Shell (VPDB) (VPDB) (temp.) (D47) temp.) color sc-7 San Simeon, CA 3 Helicidae Helix À10.91 ± 0.02 À0.66 ± 0.11 0.629 ± 0.007 29 ± 3 1.48 À1.31 À6.2 DB aspersa sc-6 San Simeon, CA 3 Helicidae Helix À11.03 ± 0.21 À0.57 ± 0.16 0.656 ± 0.009 23 ± 2 0.12 À1.22 À6.2 DB (O) aspersa sc-5 San Simeon, CA 4 Helicidae Helix À11.12 ± 0.28 À0.50 ± 0.03 0.663 ± 0.003 21 ± 1 À0.14 À1.15 À6.2 DB (Y) aspersa sc-4 Palo Alto, CA 3 Helicidae Helix À12.27 ± 0.05 1.16 ± 0.05 0.639 ± 0.015 27 ± 3 2.76 À0.30 À7.7 DB aspersa 6861 snails land in isotopes Clumped sc-11 Palo Alto, CA 3 Helicidae Helix À11.38 ± 0.06 0.40 ± 0.24 0.644 ± 0.010 26 ± 2 1.74 À1.06 À7.7 DB aspersa sc-2 Negev, Israel (1) 4 Sphincterochila À0.96 ± 0.04 3.02 ± 0.03 0.646 ± 0.004 25 ± 1 4.26 2.83 À5.1 W zonata sc-1 Negev, Israel (4) 3 Trochoidea simulate À6.44 ± 0.05 1.90 ± 0.01 0.651 ± 0.006 24 ± 1 2.88 1.70 À5.1 W + B sc-8 Negev, Israel (6) 3 Trochoidea simulate À7.24 ± 0.05 1.71 ± 0.02 0.640 ± 0.011 26 ± 2 3.23 1.52 À5.1 W + B sc-3 Haifa, Israel 4 Helicidae Helix À13.27 ± 0.05 0.72 ± 0.04 0.646 ± 0.010 25 ± 2 1.96 0.41 À1.5 DB aspersa sc-pb-1 Pennsylvania 5 Triodoesis notata À9.42 ± 0.05 À1.54 ± 0.11 0.653 ± 0.018 23 ± 4 À0.69 À3.81 À6.1 MB sc-pb-2 East Saginaw, 6 Triodoesis notata À9.69 ± 0.04 À3.64 ± 0.07 0.633 ± 0.016 28 ± 4 À1.76 À6.38 À8.9 MB MI sc-pb-3 Washington, DC 3 Triodoesis tridentata À13.61 ± 0.04 À2.93 ± 0.04 0.642 ± 0.009 26 ± 2 À1.49 À4.73 À4.5 MB sc-pb-4 Florence, AL 4 Triodoesis tridentata À11.30 ± 0.05 À1.08 ± 0.10 0.656 ± 0.012 23 ± 3 À0.34 À2.08 À6.5 MB sc-pb-5 St. Louis County, 5 Triodoesis notata À11.20 ± 0.05 À2.14 ± 0.06 0.623 ± 0.013 30 ± 3 0.31 À3.83 À6.8 MB MO sc-pb-6 Jamaica 3 Pleurodonte acuta À10.86 ± 0.01 À0.73 ± 0.04 0.607 ± 0.011 34 ± 3 2.58 0.85 À3MB sc-pb-7 Jamaica 3 Orthalicus undatus À8.34 ± 0.05 0.10 ± 0.07 0.634 ± 0.005 28 ± 1 1.92 1.68 À3W+B sc-pb-8 Stock Island, FL 3 Orthalicus reses À11.14 ± 0.07 À1.64 ± 0.06 0.623 ± 0.003 30 ± 1 0.77 0.04 À4.3 W + B (say) sc-9 Davos, 3 Helicidae Arianta À10.57 ± 0.05 À3.13 ± 0.02 0.673 ± 0.017 19 ± 4 À3.25 À5.86 À10.4 DB Switzerland (O) arbustorum sc-10 Davos, 3 Helicidae Arianta À10.49 ± 0.03 À2.48 ± 0.05 0.665 ± 0.006 21 ± 1 À2.21 À5.21 À10.4 DB Switzerland (Y) arbustorum Author's personal copy

6862 S. Zaarur et al. / Geochimica et Cosmochimica Acta 75 (2011) 6859–6869

Table 1b Climatologic data (seasonal temperature and amount of precipitation) of sample locations. Activity season was deﬁned according to a combination of warm and wet conditions. Location MAT Temperature Precipitation Activity season DJF JJA DJF JJA San Simeon, CA 17 7 26 18 12 Winter Palo Alto, CA 13 10 20 90 <3 Winter Negev, Israel 19 11 24 <30 <3 Winter Haifa, Israel 18 11 24 60 3 Winter PA 10 À3 20 63 150 Summer East Saginaw, MI 8 À5 20 45 105 Summer Washington, DC 12 0 22 63 165 Summer Florence, AL 15 1 23 84 225 Summer St. Louis County, MO 12 À3 23 60 150 Summer Jamaica 27 25 27 135 210 Year round Stock Island, FL 25 22 27 60 210 Summer Davos, Switzerland 8 À2 <15 <30 144 Summer

À20 °C (following Affek and Eiler (2006) and Huntington pression that is due to dissociation–recombination reac- et al. (2009)). This is required to guarantee no interference tions of CO2 in the ion source (Huntington et al., 2009). of hydrocarbons with the mass 47 measurement (Eiler and The extent of scale compression varies among mass spec- Schauble, 2004). Measurements were performed using a trometers as well as over time and source conditions (Af- Thermo-Finnigan MAT253 gas source isotope ratio mass fek et al., 2009; Dennis et al., in press). Scale compression spectrometer modified to simultaneously measure masses in our system was accounted for by analyzing CO2 at 44-49 in a dual inlet mode (in the Earth Systems Center equilibrium at 25 °C, through isotope exchange with for Stable Isotopic Studies at Yale University). The mass water, and verified through interlab comparison and mea- spectrometer cup configuration is identical to that used surements of carbonates having published D47 values, in Caltech for the original development of clumped iso- yielding D47 values that are within 0.01& from previous topes thermometry (Eiler and Schauble, 2004). A notable estimates (Dennis et al., in press). modification in the Yale mass spectrometer is a larger All climatological information is based on National Oce- aperture of the stainless steel capillary transferring CO2 anic and Atmospheric Administration (NOAA) NCEP/ from the dual inlet bellows to the mass spectrometer NCAR Reanalysis data, taken as long-term mean for the source, resulting in a 16 V mass 44 signal for 40 mbar of years 1949–2011 (Kalnay et al., 1996). CO2 bellows pressure. An additional modification is the replacement of the original bellows potentiometer with a 3. RESULTS Bourns 3540s enabling a more efficient operative range of the bellows. These two modifications enabled the anal- D47 values observed in snail samples range between ysis of smaller samples (3–4 mg CaCO3) than previously 0.607 ± 0.011& and 0.673 ± 0.017& (Table 1). Using the described (Huntington et al., 2009). Each measurement carbonate clumped isotopes thermometer calibration consisted of 90 cycles of a sample-standard comparison, (Ghosh et al., 2006) these values correspond to a tempera- with a signal integration time of 8 s. Isotopic measure- ture range of 34 ± 3 and 19 ± 3.5 °C(Fig. 1). This range of ments were performed in replicates of n = 3–6 per temperatures is smaller than the range of mean annual tem- specimen, considering at least three replicates as a require- peratures (between 27 and 8 °C) or the estimated range of ment for a reliable result. Reported errors are the statisti- snail activity season temperatures (between 27 and 7 °C) cal standard errors of these replicates, reflecting external in the sites from which the snail samples were collected. analytical error, with the mass spectrometry data of all As snail activity requires both warmth and moisture, we de- 90 cycles showing Gaussian distributions and standard er- fine season of activity as the warm, rainy season; namely, ror that is close to the shot noise limit (Merritt and Hayes, year-round in tropical climates, winter in Mediterranean 1994). Both the d18O and d13C are referenced to the V- climates (in which summers are too dry), and summer in PDB scale as determined using a pre-calibrated Oztech temperate climates (in which winters are too cold). The ob- 18 13 CO2 tank as a reference working gas, and verified using served d O and d C values range between À0.96& and NBS-19. D47 is defined as the excess of CO2 mass 47 signal À13.61& and +3.02& and À3.64&, respectively. No cor- over what is expected for random distribution of 13C and relation was found between shells d13C and d18O 18O among all isotopologues (Eiler et al., 2008; Hunting- (Fig. 2a), nor is there a correlation between either d13Cor 18 ton et al., 2009). Standardization is done using CO2 that d O and D47 values (Fig. 2b). is heated at 1000 °C for 2 h, resulting in randomization Intra-shell variations were examined by comparing seg- (Wang et al., 2004). A series of such heated gases having ments growing at different stages of the snail life cycle À a range of bulk compositions was used to correct for mass the upper part of the helix, reflecting material produced spectrometer effects such as non-linearity and scale com- early in the snail’s life vs. the lower part of the helix that Author's personal copy

Clumped isotopes in land snails 6863

Fig. 1. (a) Clumped isotopes values in snail shells (left axis) and temperatures calculated (right axis) based on Ghosh et al. (2006). Snail locations are ordered according to increasing mean annual temperatures (from left to right). Snails of the same species are marked by the same symbol. (b) Mean annual (diamonds) and season activity (squares) temperatures of each location.

18 13 18 13 Fig. 2. d O vs. d C (a) and D47 vs. both d O and d C (b) observed in land snail shells. In all cases no correlations are observed (linear 2 18 13 18 13 regression R values of 0.333, 0.003 and 0.003 for d O vs. d C and D47 vs. either d Oord C, respectively). formed later (specimens from San Simeon, CA, and Davos, species (Table 3), as well as two different species collected Switzerland; Table 2). No significant difference in either in the same location (Table 4), for several locations. 13 D47 or d C were observed within a specimen, with D47 values Detailed data are given in Table 1. In the Negev desert, of 0.656 ± 0.009& and 0.663 ± 0.003& (sc-5 and sc-6 from San Simeon, CA; Table 1) and 0.673 ± 0.017& and 0.665 ± 0.006& (sc-9 and sc-10 from Davos, Switzerland). Table 2 Respective d13C values are 11.03 ± 0.21& and 11.12 ± À À Intra shell comparison of D47 in different growth stages of Helicidae 0.28& (sc-6 and sc-5), and À10.57 ± 0.05& and À10.49 ± Helix aspersa from San Simeon, CA and Helicidae Arianta 0.03& (sc-9 and sc-10). Variations in d18O, when comparing arbustorum from Davos, Switzerland. Upper part of the helix the lower and upper helix, were not significant in the San represents an early growth stage of the snail (Y) and lower part of Simeon samples: À0.57 ± 0.16& and À0.50 ± 0.03& (sc-6 helix represents an older age (O). and sc-5), whereas somewhat larger variations were observed Sample D47 (&) in the Davos sample: 3.13 ± 0.02& and 2.48 ± 0.05& À À San Simeon, CA (O) 0.656 ± 0.009 (sc-9 and sc-10). All other data is reported for whole shell San Simeon, CA (Y) 0.663 ± 0.003 analysis. To examine inter- and intra-species variations, we Davos, Switzerland (O) 0.673 ± 0.017 Davos, Switzerland (Y) 0.665 ± 0.006 compared D47 values of two individual snails of the same Author's personal copy

6864 S. Zaarur et al. / Geochimica et Cosmochimica Acta 75 (2011) 6859–6869

Table 3 A comparison of the same species from the same location in the Negev, Israel, San Simeon and Palo Alto, CA.

Sample Species D47 (&) Negev, Israel (sc-1) Trochoidea simulate 0.651 ± 0.006 Negev, Israel (sc-8) Trochoidea simulate 0.640 ± 0.011 Palo Alto, CA (sc-4) Helicidae Helix aspersa 0.639 ± 0.011 Palo Alto, CA (sc-11) Helicidae Helix aspersa 0.644 ± 0.015 San Simeon, CA (sc-5 + 6) Helicidae Helix aspersa 0.660 ± 0.006 San Simeon, CA (sc-7) Helicidae Helix aspersa 0.629 ± 0.013

Table 4 information about paleo-vegetation and related hydrologi-

Comparing the D47 of different snail species from the same location. cal conditions. We can therefore infer that, with the exception of one specimen from the Negev, Israel, all the Sample Species D47 (&) analyzed snails fed almost exclusively on C plants, in Negev, Israel (sc-2) Sphincterochila zonata 0.646 ± 0.004 3 agreement with the C dominance in their environment Negev, Israel (sc-1) Trochoidea simulate 0.651 ± 0.006 3 Negev, Israel (sc-8) Trochoidea simulate 0.640 ± 0.011 (Woodward and Lomas, 2004). Sample sc-1 from The Ne- gev, Israel, had an anomalously high isotopic d13C value of Jamaica (sc-pb-6) Pleurodonte acuta 0.607 ± 0.011 À0.96&, and most likely fed on C4 or CAM plants. Jamaica (sc-pb-7) Orthalicus undatus 0.634 ± 0.007 In general, trends in oxygen isotope composition of a shell are related to local environmental conditions, namely, more 18O enriched values are characteristic of hot and dry Israel, D47 values of Sphincterochila zonata (sc-2) and two regions and more depleted values are associated with colder specimens of Trochoidea simulate (sc-1 and sc-8) did not and more humid environments. However, as noted previ- vary significantly with D47 values of 0.646 ± 0.004&, ously (Lecolle, 1985; Zanchetta et al., 2005; Yanes et al., 0.651 ± 0.006&, and 0.640 ± 0.011&, respectively. Their 2009), this correlation is not straightforward. The expected d13C values, on the other hand, were highly variable, espe- correlation with environmental conditions such as ambient cially between species, with values of À0.96 ± 0.04& (sc-2), temperature and rainfall d18O yielded reasonable estimates À6.44 ± 0.05& and À7.24 ± 0.05& (sc-1 and sc-8). d18O in some cases, but this approach did not hold in cases of values differed somewhat among species but only slightly low ambient temperatures (Lecolle, 1985). Yet, in spite of among individuals within the same species, with values of this complication, most geochemical studies assume that +3.02 ± 0.03& (sc-2), +1.90 ± 0.01& and +1.71 ± 0.02& calcification temperature is identical to ambient tempera- (sc-1 and sc-8). The comparison of Pleurodonte acuta and ture (either mean annual or snail growing season). Our Orthalicus undatus from Jamaica showed significant clumped isotopes data indicate that calcification tempera- differences in D47, with values of 0.607 ± 0.011& and tures are higher than ambient temperatures, and in cold 0.634 ± 0.005&, (sc-pb-6 and sc-pb-7), as well as in d13C environments the difference tends to be larger, providing with values of À10.86 ± 0.01& and À8.34 ± 0.05& and a likely explanation for these oxygen isotope discrepancies. in d18O, with values of À0.73 ± 0.04& and +0.10 ± 0.07&, respectively. Two Helicidae Helix aspersa (sc-4 4.2. Clumped isotopes and sc-14) individuals from Palo Alto, CA, showed no significant difference in D47 (0.639 ± 0.015& and 4.2.1. Calcification temperatures 0.644 ± 0.010&), and a similar range of d13C and d18O in- The range of calcification temperatures recorded by the tra-species variability (À12.27 ± 0.05& and À11.38 ± snails is small (19–34 °C) relative to the range of the tem- 0.06& for d13C, +1.16 ± 0.05& and +0.40 ± 0.24& for peratures associated with snail activity season in the differ- d18O). In San Simeon, CA, however, two individuals of ent locations (7–27 °C). Seasonal and diurnal variations Helicidae Helix aspersa (sc-7 and sc-5-6), showed significant could have provided an explanation to the temperature dis- variability in D47 with values of 0.629 ± 0.013& and crepancy; however, we found no correlation between either 0.660 ± 0.006&, but only small variability in d13C mean annual ambient or snail activity season temperatures (À10.91 ± 0.02& and À11.07 ± 0.25&) and d18O(À0.66 ± and temperatures obtained from clumped isotopes (Figs. 1 0.11& and À0.53 ± 0.10&), for samples sc-7 and sc-5-6, and 3a). In some locations (Israel and California) D47 re- respectively. flects temperatures that are close to the local annual mean even though snails in these regions are active almost only 4. DISCUSSION during the winter (Heller, 1993). In most other locations, the temperature recorded by the snail shells is higher than 4.1. Carbon and oxygen isotopes ambient, even when accounting for a possible bias towards shell growth in summer. Carbon isotopes in snail shells are typically interpreted Comparing the clumped isotopes temperatures to the lo- to reflect snail diet (Stott, 2002), in particular the rela- cal ambient temperatures (either mean annual or snail tive contribution of C4 vs. C3 vegetation, thus providing activity season; Fig. 3a) reveals a general trend in which Author's personal copy

Clumped isotopes in land snails 6865

Fig. 3. (a) A comparison of measured clumped isotopes temperatures in land snail shells to local mean annual and snail activity season temperatures. (b) The offset of clumped isotopes derived snail calcification temperatures from activity season temperature as compared to the local season activity temperature. Note that the linear fit is a convenient approximation but is not a physiology-based function and is therefore valid for interpolation but not extrapolation. higher environmental temperatures correlate with higher regions snail calcification temperatures may be lower than growth temperatures. However, the absolute calcification ambient. The observed D47 values thus reflect the tempera- temperatures in our samples are always higher than local ture at the site of calcification, which is likely to be close to temperatures, with calcification temperatures being close snail body temperature during the time of snail activity. to uniform across most locations. The observed tempera- The observation that calcification and ambient temperature offset between clumped isotopes temperatures and lo- tures differ from each other has implications beyond snail cal snail activity season temperatures is higher where eco-physiological aspects, and addresses specifically the environmental temperatures are lower (Fig. 3b). This trend use of snail shells for paleo-environmental reconstruction. is consistent with past observations in which environmental This calcification temperature reflects the conditions re- temperatures could plausibly explain the observed shell corded by shell d18O and is therefore the temperature rele- d18O in warm regions, but not in colder, high elevation sites vant for estimating d18O of the associated water, typically (Lecolle, 1985; Yanes et al., 2009). interpreted as a paleo-hydrology indicator. Clumped iso- Although generally ignored in geochemical studies, it topes in land snails therefore serve as a way to correctly re- has been shown that coloration and morphology of the solve the temperature and water signatures in shell d18O, shell, as well as behavioral lifestyle adaptation may signifi- required to reconstruct snail body water d18O as a proxy cantly affect snail body temperature (Heath, 1975; for hydrological parameters such as rainfall and relative Dittbrenner et al., 2009), leading to snail body tempera- humidity. tures, and hence calcification temperatures, that are typically warmer than ambient. Dark colored shells may 4.2.2. Shell growth under equilibrium conditions increase snail body temperature by up to 12 °C above envi- Whereas the notion of calcification temperature that is ronmental temperature by exposure to direct sunlight, as higher than ambient temperature is consistent with biolog- the snail basks in the sun (Heath, 1975). At the same time, ical evidence for elevated snail body temperature, an alter- thick white shells and a lifestyle that involves being away native potential explanation to the D47 observations may be from the warm soil surface (by climbing on vegetation) aragonite formation under non-equilibrium conditions. Iso- may moderate snail body temperature in regions where soil topic dis-equilibrium in D47 has been observed in speleo- and daytime air temperatures exceed a critical value (Heller, them carbonate (Affek et al., 2008; Dae¨ron et al., 2011) 1993). Shell color, morphology, and behavior may thus and is associated with CO2 degassing out of solution provide a specific adaptation that allows the snail some (Guo, 2008). control of its body temperature. Indeed, in our measure- Precipitation under conditions of isotopic equilibrium is ments, the largest offset between clumped isotopes temper- a key assumption in interpretation of carbonate isotopic atures and ambient temperatures was observed in the dark composition based on an empirical d18O–T calibration rela- brown colored shells. The smallest difference was observed tionship (Epstein et al., 1953; Grossman and Ku, 1986; Kim in the lighter, medium-gray shells. In view of this, it might and O’Neil, 1997; Kim et al., 2007), although potential be surprising that white shells of snails from the Negev, Is- deviation from equilibrium in these d18O–T calibrations is rael, show a larger temperature difference. This, however, often argued (e.g., Coplen, 2007; Dietzel et al., 2009). For can result from other morphological or behavioral controls clumped isotopes, a D47–T relationship developed by on body temperature, such as longer hours of exposure to Ghosh et al. (2006) followed calcite precipitation proce- the sun during the winter activity period. Though not ob- dures that were based on the Kim and O’Neil (1997) ap- served in our dataset, we expect that in especially warm proach and is therefore assumed to reflect a nominal Author's personal copy

6866 S. Zaarur et al. / Geochimica et Cosmochimica Acta 75 (2011) 6859–6869 equilibrium relationship. This assumption is based primar- depend on a combination of biological and environmental ily on the observed conformity of a large variety of biogenic conditions (Heath, 1975; Heller, 1993). For example, a snail carbonates to that calibration (summarized by Eiler, 2007; growing in warm, dry conditions, such as the Negev in Is- Tripati et al., 2010), including mostly marine organisms rael, is likely not to be active during peak warmth and to but also some terrestrial carbonates such as teeth (Eagle prefer activity in winter. An additional adaptation would et al., 2010). The conformity of biogenic carbonates that be the thick white shell of Sphincterochita zonata that pre- are either calcite, aragonite, or apatite to the same D47–T vents excessive heating of snail body (Heller, 1993). In that relationship suggests that there is no clumped isotopes frac- view, it is not surprising to see a difference in D47 values be- tionation associated with the mineral formation, so that tween the two differently colored snails from warm, humid isotopic equilibrium distribution of 13C–18O bonds is deter- conditions in Jamaica, with the species of a darker color mined in the solution (Guo, 2008). A notable exception (Pleurodonte acuta) showing higher calcification tempera- from the nominal equilibrium conformity are speleothem ture (Table 4). Though not direct evidence for equilibrium carbonates that significantly deviate from that relationship carbonate formation, these observations are in support of (Affek et al., 2008; Dae¨ron et al., 2011; Wainer et al., 2011) a biological adaptation of calcification temperature, rather as a result of kinetic isotope effects associated with CO2 than a non-equilibrium isotope effect as the main control of degassing and hydration and hydroxylation reactions of D47 values. It is encouraging, however, to note that snails of CO2 (Guo, 2008). A priori we expect that clumped isotopes similar color and shell thickness that grow under similar cli- in land snails would follow the general biogenic D47–T rela- matic conditions, such as Israel and California (except for tionship, assuming that snail body solution is at isotopic one individual in San Simeon, CA, discussed below) show equilibrium. Deviation from equilibrium, however, may similar D47 values, consistent with isotopic equilibrium be a potential alternative explanation to the observed low- (Table 3). er-than-expected D47 values. We therefore conducted a Contrary to these two samples, the two specimens of number of tests, as described below, to help us assess pre- Helicidae Helix aspersa from San Simeon, CA, show D47 cipitation conditions. values that are significantly different (0.660 ± 0.006 and Kinetic isotope effects are expected to be associated with 0.629 ± 0.013). It is possible that the specimen that re- variations in mineral growth rates and are likely to be recorded lower calcification temperatures lived in a shady, flected as heterogeneity in D47 values within a shell. In prin- more humid micro-environment (such as a shady garden ciple, this could be tested in snails that grow year round or lawn) while the other lived in a sunnier micro-environ- with growth rates that vary seasonally. Unfortunately, the ment or that one of them happened to capture anomalous relatively large sample size required for clumped isotopes cool or warm climatic years. Even though these samples analysis (15 mg for 3–4 replicates) does not allow such were collected in the same location, the difference in their high resolution sampling within one shell. Furthermore, actual living micro-environments could affect their bodies seasonal variations in d13C and d18O may merely reflect and hence calcification temperatures. It seems very likely variations in diet and body water as opposed to dis-equilib- that these snails, collected as shell fragments of snails living rium. We therefore focus on low resolution phenological in a garden, did not grow in a natural environment and comparison of shell produced early in the snail life cycle hence would not necessarily reflect regional conditions. (young) vs. shell produced by a mature snail, irrespective of seasonality. No significant young vs. old difference was 4.3. Rain water reconstruction observed in two specimens growing under different conditions (Helicidae Helix aspersa from San Simeon, CA and Early studies of snail shell carbonates (e.g., Yapp, 1979) Helicidae Arianta arbustorum from Davos, Switzerland; Ta- attempted to correlate shell and rain-water d18O values bles 1 and 2), in spite of observed variations in d13C and using mean annual temperatures. As no direct relationship d18O values, consistent with shell aragonite forming under was found, later studies used the temperatures and rainfall isotopic equilibrium. d18O associated with times of snail activity (Leng et al., Deviation from equilibrium will affect not only D47 but 1998; Balakrishnan et al., 2005). We combine clumped iso- also d18O and d13C so that under kinetically controlled con- topes temperatures with shell d18O to calculate d18O of snail ditions they are expected to be correlated with each other. body water (using the Grossman and Ku (1986) calibration No such correlation was observed either within individual as corrected by Kobashi and Grossman (2003)). This is then shells or among all snail samples (Fig. 2). Although in itself compared to estimated d18O of local precipitation (based on this is not a proof for isotopic equilibrium, as d13C and the mapping model of Bowen and Revenaugh (2003) as gi- d18O may vary due to uncharacterized external parameters, ven in wateriso.eas.purdue.edu/waterisotopes/; Fig. 4). such as snail diet, the absence of correlation does not sup- Since snail growth temperatures are generally higher than port kinetic isotope effects as the main mechanism for D47 mean annual temperatures, the calculated isotopic compo- offset from ambient temperatures. sitions of snail body water when using the clumped isotopes There is no a priori reason to expect that different species temperatures are more enriched than those obtained using at one location would show the same D47 value. Nor is it ex- environmental temperatures (by 0.2–4.1&). These water pected that individuals of the same species from different values are slightly better correlated with rainfall d18O(R2 climatic regimes would show the same body warming above of 0.784 vs. 0.735), highlighting the link between rain and ambient temperatures. Instead, the offset from the D47 ex- body water that is modified, however, by enhanced evapo- pected for growing season ambient temperatures would rative enrichment in dry regions (Fig. 4). Author's personal copy

Clumped isotopes in land snails 6867

indicates that knowing the actual calcification temperature is necessary in order to estimate water composition and that this might vary among regions and climatic conditions, and be affected by species-specific parameters such as color and shell morphology. Using land snails for paleo-hydrology reconstruction would further benefit from an understanding of the link between rainfall and body water d18O. This may be facilitated by focusing snail-based paleo-hydrological analysis on specific regions and combining it with a thorough regional-scale characterization of the relationship between modern-day rain and snail d18O. Although it is possible that the offset between calcification temperature and ambient temperatures is in itself climate-related (and therefore region-specific), it is more likely that it reflects species-specific biological and behav- Fig. 4. Oxygen isotope composition of the snail body water, ioral adaptations. These biological adaptations are likely calculated from clumped isotopes temperatures and shell d18O, to vary in response to prolonged exposure to different envi- compared to d18O of annual local precipitation. Regions are divided to humid (RH > 70%) and dry (RH < 70%). Precipitation ronmental conditions associated with climate states that are d18O is based on Bowen and Revenaugh (2003). RH data from radically different from the Holocene, such as those preva- NCEP/NCAR reanalysis (Kalnay et al., 1996). lent over the glacial–interglacial cycles. In other words, although in a colder climate a snail might not be able to Snails obtain water through the integument of the foot warm its body to the same temperature possible under (Prior, 1985; Heller, 1993), so that the source of body water modern-day conditions, it does not necessarily imply a con- is mostly rain, dew, and soil surface water, and a possible stant ambient-to-body temperature offset, as the species is small contribution from digested vegetation, with both rain able to adapt over time to the colder climate. As such, a re- and dew reflecting condensed local vapor (Welp et al., gional analysis of modern snails may not be sufficient to de- 2008). Snail body water would therefore reflect the local velop a calcification-to-ambient temperature transfer rainfall, with d18O modified by soil surface evaporation, function and the independent information provided by the extent of which depends mostly on the local relative clumped isotopes would be required to determine paleo-cal- humidity (RH). Snail shells reflect the rainfall occurring cification temperature. The potential climatic information stored in snail shells is therefore related to hydrological during the snails’ growing season, with most snails being 18 dormant during the dry or cold seasons. parameters that can be derived from d O, using the calci- The effect of evaporation may be characterized by sepa- fication temperature derived from D47. rately examining body water in snails that grow in humid vs. dry conditions, with low RH resulting in more enriched 4.4. Implications for previous studies – revisiting Yapp (1979) body water (Fig. 4). Our data clearly divides into two groups following the local RH (high RH is >75%, low To illustrate the potential use of snail isotopic analysis RH is <70%; Fig. 4). In both groups, snail body water as a paleo-hydrological tool, we used our clumped-isotopes shows clear 18O enrichment relative to local rainfall, with based calcification temperatures to re-examine one of the enrichment being more pronounced at low RH locations. early published datasets. Yapp (1979) performed a regional Within the high RH locations there is a weak but noticeable d18O study, comparing different modern species from cli- trend with temperature. Precipitation d18O values in war- matically diverse regions (Utah, Wisconsin, South mer regions are typically higher, but show a lower effect Carolina, Washington, Ohio, Louisiana, Mexico, and Nor- of evaporative enrichment in snail body water; instead, way) with estimated environmental temperatures ranging 18 the d O difference between precipitation and snail body 13–22 °C and RH ranging 50–80%. Given that no D47 data water reflects the temperature dependence of equilibrium is available for that dataset, we estimate calcification tem- fractionation factor in evaporation. perature based on the offset observed in our data from The two San Simeon snails discussed above as recording ambient temperature (both MAT given in the original study calcification temperatures that are significantly different and snail activity season temperatures average based on from each other are split between the two RH groups. NCAR reanalysis data; Fig. 3b). It is important to note that The higher temperature one falls within the low RH group, these calcification temperatures are only a first estimate and close to the samples from adjacent Palo Alto, whereas the should not replace direct measurements, but may still serve other specimen is an outlier, and plots within the high as an illustrative example. These temperatures are then RH group, consistent with our assumption that this indi- combined with shell d18O values to reconstruct the respec- vidual represents anomalous conditions, either a human- tive snail body water d18O (see SI for the complete dataset). modified (shaded and irrigated) micro-environment or an Since calcification temperatures are always higher than unusual year in terms of both temperature and moisture environmental temperatures, the resulting snail body water availability. is more enriched than the original estimates, thus revealing Body water d18O can thus be used as an indicator of the evaporative enrichment of surface water. Also here, dry variations in rainfall d18O and local evaporation. Our data regions (Sevier County, Utah; Lower Garnd Coulee, Author's personal copy

6868 S. Zaarur et al. / Geochimica et Cosmochimica Acta 75 (2011) 6859–6869

Washington; and Taos County, New Mexico) reflect large Balakrishnan M. and Yapp C. J. (2004) Flux balance models for water enrichment. The limited temperature range (20– the oxygen and carbon isotope compositions of land snail 22 °C) of the dataset, however, does not allow testing for shells. Geochim. Cosmochim. Acta 68, 2007–2024. Balakrishnan M., Yapp C. J., Theler J. L., Carter B. J. and enrichment trend with temperature. Using mean annual 13 12 environmental temperatures, Yapp (1979) showed that the Wyckoff D. G. (2005) Environmental significance of C/ C and 18O/16O ratios of modern land-snail shells from the degree of isotopic enrichment in the snail is negatively cor- southern great plains of North America. Quatern. Res. 63, related with local RH. Our re-analysis of his data does not 15–30. change this basic conclusion, but emphasizes the role of lo- Barker G. M. (2001) The biology of terrestrial molluscs, i–xiv, cal evaporation. CABI Publishing, New York. pp. 1–558. Bowen G. J. and Revenaugh J. (2003) Interpolating the isotopic 4.5. Paleo-climate implications composition of modern meteoric precipitation. Water Resour. Res. 39(10), 1299. Whereas our observations have interesting implications Coplen T. B. (2007) Calibration of the calcite-water oxygen-isotope for the eco-physiological behavior of land snails, they also geothermometer at Devils Hole, Nevada, a natural laboratory. suggest that the use of snail shells for paleo-climatic recon- Geochim. Cosmochim. Acta 71, 3948–3957. Dae¨ron M., Guo W., Eiler J., Genty D., Balmart D., Boch R., struction on land is not as straightforward as may have Drysdale R., Maire R., Wainer K. and Zanchetta G. (2011) been hoped. Clumped isotopes temperatures in land snail 13C18O clumping in speleothems: observations from natural shells cannot be used to directly reconstruct air tempera- caves and precipitation experiments. Geochim. Cosmochim. tures, but can be useful in a paleo-hydrological context as Acta 75, 3303–3317. they provide the calcification temperatures required to ex- Dennis K. J., Affek H. P., Passey B. H., Schrag D. P. and Eiler J. tract rainwater isotopic signals out of shell oxygen isotopic M. (in press) Defining an absolute reference frame for composition. Snail shell isotopic composition records cli- ‘clumped’ isotope studies of CO2. Geochim. Cosmochim. Acta, matic conditions, but these are modified by species-specific doi:10.1016/j.gca.2011.09.025. calcification temperatures and region-specific evaporative Dietzel M., Tang J. W., Leis A. and Kohler S. J. (2009) Oxygen enrichment at the soil surface. Hence, land snail-based pa- isotopic fractionation during inorganic calcite precipitation – effects of temperature, precipitation rate and pH. Chem. Geol. leo-hydrology requires a comparison with modern oxygen 268, 107–115. and clumped isotopes analysis, preferably of similar snail Dittbrenner N., Lazzara R., Ko¨hler H., Mazzia C., Gapowiez Y. species, to provide a baseline for regional evaporation and Triebskorn R. (2009) Heat tolerance in Mediterranean land and its effect on snail body water. snails: histopathology after exposure to different temperature regimes. J. Mollus. Stud. 75, 9–18. ACKNOWLEDGMENTS Eagle R. A., Schauble E. A., Tripati A. K., Tutken T., Hulbert R. C. and Eiler J. M. (2010) Body temperatures of modern and We thank Eric Lazo-Wasem and Lourdes Rojas from the Yale extinct vertebrates from 13C–18O bond abundances in bioapa- Peabody Museum of Natural History for sharing samples and tite. Proc. Natl. Acad. Sci. USA 107, 10377–10382. Adam Baldinger from the Harvard University Museum of Com- Eiler J. M. (2007) “Clumped-isotope” geochemistry – the study of parative Zoology for assistance in snail species identification. We naturally-occurring, multiply-substituted isotopologues. Earth thank Sharon Reinhart, Dominic Colosi and the Earth System Planet. Sci. Lett. 262, 309–327. Center for Stable Isotope Studies of the Yale Institute for Bio- Eiler J. M., Affek H., Daeron M., Ferry J., Guo W. F., Huntington spheric Studies. We thank K. Huntington, M. Dae¨ron and C. Yapp K., Thiagarajan N. and Tripati A. (2008) Carbonate ‘clumped for their thorough review and helpful comments. The work was isotope’ thermometry: a status report. Geochim. Cosmochim. supported by the National Science Foundation Grant NSF-EAR- Acta 72, A239. 0842482 to H.P.A. Eiler J. M. and Schauble E. (2004) 18O13C16O in Earth’s atmosphere. Geochim. Cosmochim. Acta 68, 4767–4777. Epstein S., Buchsbaum R., Lowenstam H. A. and Urey H. C. APPENDIX A. SUPPLEMENTARY DATA (1953) Revised carbonate-water isotopic temperature scale. Geol. Soc. Am. Bull. 64, 1315–1325. Ghosh P., Adkins J., Affek H., Balta B., Guo W. F., Schauble E. Supplementary data associated with this article can be A., Schrag D. and Eller J. M. (2006) 13C18O bonds in carbonate found, in the online version, at doi:10.1016/j.gca.2011. minerals: a new kind of paleothermometer. Geochim. Cosmo- 08.044. chim. Acta 70, 1439–1456. Goodfriend G. A. (1990) Rainfall in the Negev desert during the REFERENCES middle Holocene, based in 13C of organic-matter in land snail shells. Quatern. Res. 34, 186–197. Affek H. P., Bar-Matthews M., Ayalon A., Matthews A. and Eiler Goodfriend G. A. (1992) The use of land snail shells in J. M. (2008) Glacial/interglacial temperature variations in paleoenvironmental reconstruction. Quatern. Sci. Rev. 11, Soreq cave speleothems as recorded by ‘clumped isotope’ 665–685. thermometry. Geochim. Cosmochim. Acta 72, 5351–5360. Goodfriend G. A. and Ellis G. L. (2002) Stable carbon and oxygen

Affek H. P. and Eiler J. M. (2006) Abundance of mass 47 CO2 in isotopic variations in modern Rabdotus land snail shells in the urban air, car exhaust, and human breath. Geochim. Cosmo- southern Great Plains, USA, and their relation to environment. chim. Acta 70, 1–12. Geochim. Cosmochim. Acta 66, 1987–2002. Affek H. P., Zaarur S. and Douglas P. M. J. (2009) Mass Goodfriend G. A. and Magaritz M. (1987) Carbon and oxygen spectrometric effects on ‘clumped isotopes’ calibration. Geo- isotope composition of shell carbonate of desert land snails. chim. Cosmochim. Acta 73, A15. Earth Planet. Sci. Lett. 86, 377–388. Author's personal copy

Clumped isotopes in land snails 6869

Goodfriend G. A., Magaritz M. and Gat J. R. (1989) Stable isotope Magaritz M. and Heller J. (1980) A desert migration indicator – composition of land snail body-water and its relation to oxygen isotopic composition of land snail shells. Palaeogeogr. environmental water and shell carbonate. Geochim. Cosmochim. Palaeoclimatol. Palaeoecol. 32, 153–162. Acta 53, 3215–3221. Merritt D. A. and Hayes J. M. (1994) Factor controlling precision Grossman E. L. and Ku T. L. (1986) Oxygen and carbon isotope and accuracy in isotope-ratio mass-spectrometry. Anal. Chem. fractionation in biogenic aragonite – temperature effects. Chem. 66, 2336–2347. Geol. 59, 59–74. Metref S., Rousseau D. D., Bentaleb I., Labonne M. and Vianey- Guo W. (2008) Carbonate clumped isotope thermometry: applica- Liaud M. (2003) Study of the diet effect on delta 13C of shell tion to carbonaceous chondrites and effects of kinetic isotope carbonate of the land snail Helix aspersa in experimental fractionation. Ph.D. thesis, California Institute of Technology. conditions. Earth Planet. Sci. Lett. 211, 381–393. Heath D. J. (1975) Colour, sunlight and internal temperatures in Prior D. J. (1985) Water-regulatory behavior in terrestrial gastro- the land-snail Cepaea nemoralis (L.). Oecologia 19, 29–38. pods. Biol. Rev. Camb. Philos. Soc. 60, 403–424. Heller J. (1993) Land snails of the land of Israel: natural history Schauble E. A., Ghosh P. and Eiler J. M. (2006) Preferential and a field guide, Ministry of Defence Publishing, Isreal (in formation of 13C–18O bonds in carbonate minerals, estimated Hebrew). pp. 1–271. using first-principles lattice dynamics. Geochim. Cosmochim. Huntington K. W., Eiler J. M., Affek H. P., Guo W., Bonifacie M., Acta 70, 2510–2529. Yeung L. Y., Thiagarajan N., Passey B., Tripati A., Daeron M. Stott L. D. (2002) The influence of diet on the delta 13C of shell

and Came R. (2009) Methods and limitations of ‘clumped’ CO2 carbon in the pulmonate snail Helix aspersa. Earth Planet. Sci. isotope (D47) analysis by gas-source isotope ratio mass spec- Lett. 195, 249–259. trometry. J. Mass Spectrom. 44, 1318–1329. Tripati A. K., Eagle R. A., Thiagarajan N., Gagnon A. C., Bauch Kalnay E., Kanamitsu M., Kistler R., Collins W., Deaven D., H., Halloran P. R. and Eiler J. M. (2010) 13C18O isotope Gandin L., Iredell M., Saha S., White G., Woollen J., Zhu Y., signatures and ‘clumped isotope’ thermometry in foraminifera Chelliah M., Ebisuzaki W., Higgins W., Janowiak J., Mo K. C., and coccoliths. Geochim. Cosmochim. Acta 74, 5697–5717. Ropelewski C., Wang J., Leetmaa A., Reynolds R., Jenne R. Wainer K., Genty D., Blamart D., Dae¨ron M., Bar-Matthews M., and Joseph D. (1996) The NCEP/NCAR 40-year reanalysis Vonhof H., Dublyansky Y., Pons-Branchu E., Thomas L., van project. Bull. Am. Meteorol. Soc. 77, 437–471. Caleteren P., Quinif Y. and Cailon N. (2011) Speleothem record Kehrwald N. M., McCoy W. D., Thibeault J., Burns S. J. and of the last 180 ka in Villars cave (SW France): investigation of a Oches E. A. (2010) Paleoclimatic implications of the spatial large d18O shift between MIS6 and MIS5. Quatern. Sci. Rev. 30, patterns of modern and LGM European land-snail shell d18O. 130–146. Quatern. Res. 74, 166–176. Wang Z. G., Schauble E. A. and Eiler J. M. (2004) Equilibrium Kim S. T., O’Neil J. R., Hillaire-Marcel C. and Mucci A. (2007) thermodynamics of multiply substituted isotopologues of Oxygen isotope fractionation between synthetic aragonite and molecular gases. Geochim. Cosmochim. Acta 68, 4779–4797. water: influence of temperature and Mg2+ concentration. Welp L. R., Lee X., Kim K., Griffis T. J., Billmark K. A. and Baker Geochim. Cosmochim. Acta 71, 4704–4715. J. M. (2008) d18O of water vapour, evapotranspiration and the Kim S. T. and O’Neil J. R. (1997) Equilibrium and nonequilibrium sites of leaf water evaporation in a soybean canopy. Plant Cell oxygen isotope effects in synthetic carbonates. Geochim. Cos- Environ. 31, 1214–1228. mochim. Acta 61, 3461–3475. Woodward F. I. and Lomas M. R. (2004) Vegetation dynamics – Kobashi T. and Grossman E. L. (2003) The oxygen isotopic record simulating responses to climatic change. Biol. Rev. 79, 643–670. of seasonality in Conus shells and its application to under- Yanes Y., Delgado A., Castillo C., Alonso M. R., Ibanez M., De la standing late middle Eocene (38 Ma) climate. Paleontol. Res. 7, Nuez J. and Kowalewski M. (2008) Stable isotope (d18O, d13C 343–355. and dD) signatures of recent terrestrial communities from a Lecolle P. (1985) The oxygen isotope composition of landsnail low-latitude, oceanic setting: endemic land snails, plants, rain, shells as a climatic indicator – application to hydrology and and carbonate sediments from the eastern Canary Islands. paleoclimatology. Chem. Geol. 58, 157–181. Chem. Geol. 249, 377–392. Leng M. J., Heaton T. H. E., Lamb H. F. and Naggs F. (1998) Yanes Y., Romanek C. S., Delgado A., Brant H. A., Noakes J. E., Carbon and oxygen isotope variations within the shell of an Alonso M. R. and Ibanez M. (2009) Oxygen and carbon stable African land snail (Limicolaria kambeul chudeaui Germanin): a isotopes of modern land snail shells as environmental indicators high-resolution record of climate seasonality? Holocene 8, 407– from a low-latitude oceanic island. Geochim. Cosmochim. Acta 412. 73, 4077–4099. Leone G., Bonadonna F. and Zanchetta G. (2000) Stable isotope Yapp C. J. (1979) Oxygen and carbon isotope measurements of record in mollusca and pedogenic carbonate from Late Pliocene land snail shell carbonate. Geochim. Cosmochim. Acta 43, 629– soils of Central Italy. Palaeogeogr. Palaeoclimatol. Palaeoecol. 635. 163, 115–131. Zanchetta G., Leone G., Fallick A. E. and Bonadonna F. P. (2005) Li G. J., Sheng X. F., Chen J., Yang J. D. and Chen Y. (2007) Oxygen isotope composition of living land snail shells: data Oxygen-isotope record of paleorainwater in authigenic carbon- from Italy. Palaeogeogr. Palaeoclimatol. Palaeoecol. 223, 20–33. ates of Chinese loess–paleosol sequences and its paleoclimatic significance. Palaeogeogr. Palaeoclimatol. Palaeoecol. 245, 551– Associate editor: Edwin A. Schauble 559.