Calibration of the stable isotope composition and body size of the arid-dwelling land snail Sphincterochila candidissima,a climatic archive abundant in Mediterranean archaeological deposits

Abstract

Land snails recovered from archeological deposits may be used to deduce climatic conditions during prehistoric occupation because their aragonitic skeletons are usually well-preserved and document valuable climatic information in the form of isotope codes. Since the snail Sphincterochila candidissima is common in archeological sites along the western Mediterranean but has been minimally investigated, the present work examines the relationship between the oxygen (δ¹⁸O) and carbon (δ¹³C) stable isotopes of modern specimens and relevant meteorological data. Individuals of Sphincterochila were regularly live-collected throughout 1 year (from September 2013 to October 2014) in Tarragona, NE Spain, and the δ¹⁸O and δ¹³C values were measured at the shell lip or aperture, which depicts the last growth episode closest to specimens’ collection date. Shell margin δ¹⁸O values varied from +3.1‰ in September to −0.8‰ in April. Average shell margin δ¹⁸O values per collection date positively correlated with monthly averaged rainwater δ¹⁸O, and negatively correlated with monthly averaged maximum relative humidity, while other variables (i.e. temperature and precipitation amount) did not explain monthly–submonthly isotopic variations in the shell lip. Shell margin δ¹³C values ranged from −11‰ in September to −4.8‰ in March, pointing to detectable variations in snails’ diet between seasons. Interestingly, snail body size positively correlated with shell margin δ¹⁸O and δ¹³C values, suggesting that larger snail size is reached under drier conditions. This work shows that S. candidissima from the western Mediterranean is a valid paleoprecipitation δ¹⁸O and/or paleohumidity proxy in the region at the monthly–submonthly scale. Although shell margin δ¹⁸O recorded fall and spring climate conditions only, limiting therefore their use to estimate the season of harvest, snails that grow year-round (e.g. at subtropical–tropical regions) may be used usefully for such analyses.

Keywords

land snails last growth episode seasonal climate shell middens Spain stable isotopes

Introduction

Large (edible) land snails have been heavily exploited as a food resource by humans since the late Paleolithic. This activity is manifested in numerous archeological sites across the western Mediterranean region, where land snail shells are well-preserved and often concentrated in high densities (Bosch et al., 2015; Fernández-Lopez-de-Pablo et al., 2014; Hill et al., 2015; Lubell, 2004a, 2004b). Apart from subsistence strategies and food handling processes, the shells from these ‘prehistoric meals’ can be used to gain insight into the environmental conditions or climatic context during ancient human occupation (Balakrishnan et al., 2005; Colonese et al., 2007, 2010a, 2010b, 2011, 2013a, 2013b; Paul and Mauldin, 2013; Prendergast et al., 2016a; Stevens et al., 2012; Yanes et al., 2011, 2012, 2013a, 2013b, 2014). Land snail shells are entirely made of aragonite, and the oxygen (δ¹⁸O) and carbon (δ¹³C) stable isotope composition of the shell record some aspects of the environment. For example, shell δ¹⁸O values appear to track at least four atmospheric variables during calcification including rainwater δ¹⁸O, water vapor δ¹⁸O, relative humidity (RH), and temperature (Balakrishnan and Yapp, 2004). The shell δ¹³C values seem to primarily represent the consumed and assimilated plant δ¹³C values (Metref et al., 2003; Stott, 2002), possibly in combination with some ingestion of limestone (Balakrishnan and Yapp, 2004; Goodfriend et al., 1999; Liu et al., 2007; Prendergast et al., in press; Yanes et al., 2008; Zhang et al., 2014), especially in the case of large body size snails with thick shells. Considering that (1) multiple environmental factors operate jointly, often difficult to discriminate, and (2) not all land snail species behave equally, the interpretation of the stable isotope composition of snail shells remains rather challenging. The oxygen isotope composition of the snail shell is therefore particularly difficult to interpret. One way to overcome this challenge is to conduct calibration studies of the target snail species using living specimens collected from localities with well-constrained climate.

Land snail species of the genus Sphincterochila have been recovered in outstanding quantities from a large number of western and central Mediterranean archeological sites (Aparicio et al., 2015; Fernández-Lopez-de-Pablo and Gabriel, 2016; Fernández-Lopez-de-Pablo et al., 2011; Hutterer et al., 2014; Mannino and Mazzanti, 2013; Yanes et al., 2013a). The archeological shells are almost always exceptionally preserved and due to their large and thick shells, Sphincterochila is an excellent candidate for both annual-average and seasonal paleoclimatic inferences. However, studies that have examined the environmental meaning of the isotopic composition of this species are minimal (Yanes et al., 2013a), and calibration studies using living specimens to evaluate the isotopic variability of Sphincterochila candidissima at monthly–submonthly resolution remain to be undertaken.

Archeologists working on anthropogenic shell assemblages often attempt to infer some features of the subsistence strategies followed by the prehistoric human groups that occupied the target locality. In marine shells, scientists have used the δ¹⁸O values of the last growth episode, which principally depicts the water temperature closest to the organism’s collection date, to identify the season of shellfish collection and therefore, the season of site occupation (Andrus, 2011; Andrus and Thompson, 2012; Colonese et al., 2009; Hallman et al., 2009; Mannino et al., 2007; Prendergast et al., 2016b; Surge and Barrett, 2012). This type of research has never been conducted for land snail shell assemblages, in part because it has never been evaluated if adult snail shells record in the last growth episode the climatic conditions that dominate when snails are being harvested. Thus, it is as yet unexplored if we can objectively use the isotopic composition of the land snail shell margin to infer the approximate season of land snail collection.

This study investigates the seasonal variability of the oxygen and carbon isotope values of living specimens of S. candidissima collected in Tarragona (NE Spain) throughout a year. We evaluate whether or not the target species track climatic variations at the monthly–submonthly resolution for the last growth episode, and contrast isotopic data with relevant climatic variables accessible from local meteorological stations. This research is the first to investigate the potential use of land snail shell margin isotope data for estimating the approximate season of snail collection.

Material and methods

Sphincterochila candidissima

The target species belongs to the family Sphincterochilidae, which is a widely spread group in the western Mediterranean basin. All Sphincterochilidae species are adapted to arid–semiarid conditions. Moreno-Rueda (2007, 2008, 2014) and Moreno-Rueda and Collantes-Martin (2007) conducted field studies of S. candidissima in Sierra Elvira (Granada, Southern Spain). His field observations suggest that this species hibernates during the winter months to avoid cold temperatures by burying themselves under the soil surface. Moreover, the species tends to aestivate during the summer months to elude hot soil temperatures by climbing onto plants. Main activity period for S. candidissima was observed to be in the spring. This species is also capable of surviving extremely arid conditions by slowing down the rate of water loss through evaporation. Sphincterochila species exhibit a lifespan that ranges from 1 to several years. Growth rates likely vary across seasons, depending on moisture and temperature levels. Sphincterochila is broadly considered a generalized herbivore abundant in limestone areas.

Climate of Tarragona

Tarragona, located in the coast of northeast Spain at 41°06′N, 1°14′E and 68 m a.s.l. (Figure 1), has a Mediterranean climate, with mild cool winters and warm summers. Climate data were adapted from the Spanish Government Meteorological Agency website (http://www.aemet.es). Average annual air temperature is ~16°C, ranging from 9.0°C in January to 24.6°C in August (Figure 2). The area receives about 500 mm of annual precipitation. Average air temperature and precipitation during fall (September to November) and spring (March to May) is ~16°C and ~324 mm, respectively. Average annual RH is ~68%, ranging from 63% in July to 73% in October. The maximum RH value in the fall and spring seasons is ~90%, on average. Accordingly, fall and spring seasons, when Sphincterochila seems to be primarily active, tend to be the wettest seasons while winter and summer, when snails may hibernate or aestivate, are relatively dry.

Figure 1.

Map indicating the geographical location of Tarragona, NE Spain (open quadrat) where snails were live-collected between 2013 and 2014. A photograph of the target land snail species (Sphincterochila candidissima) is also shown above legend. Note that some relevant late Paleolithic and Mesolithic archeological sites rich in shells of S. candidissima are abundant in eastern Spain (filled circles).

Figure 2.

Current climate of Tarragona, NE Spain: (a) seasonal variation of rainwater oxygen isotope values and air temperature; (b) seasonal variation of the precipitation amount and average relative humidity values. Shaded areas depict the months of the year when snails are primarily active and grow shell (personal field observations, 2013–2014). Climate data were adapted from the Spanish Government Meteorological Agency website (http://www.aemet.es). Rainwater oxygen isotope values were taken from Araguas-Araguas and Diaz-Teijeiro (2005) and the International Atomic Energy Agency website (www.iaea.org).

Rainwater oxygen isotope values in the nearby coastal city of Barcelona were adopted from Araguas-Araguas and Diaz-Teijeiro (2005) and data available at the International Atomic Energy Agency website (www.iaea.org). The average annual rainwater δ¹⁸O value is ~ −5.0 ± 1.0‰ (SMOW) (1-sigma), varying between −7.0 ± 2.5‰ (SMOW) in November and −2.4 ± 2.5‰ (SMOW) in June. Rainwater δ¹⁸O values during fall and spring seasons are ~ −5.0‰ (SMOW), on average (Figure 2). The vegetation in the region is dominated by C₃-type plants. The most abundant plant species were Mediterranean shrubs (i.e. Rosmarinus officinalis, Quercus coccifera, Pistacia lentiscus) and tree forms (i.e. Pinus halepensis, Ceratonia siliqua) (personal field observations, 2013–2014).

Field sample collection

For this project, we planned a sampling strategy that has been commonly employed for marine gastropods (e.g. Colonese et al., 2009; Prendergast et al., 2013) but has never been conducted for land snails in a systematic manner (but see Goodfriend et al., 1989). A total of 44 adult specimens of S. candidissima were live-collected nine times throughout 1 year (from September 2013 to October 2014) near the city of Tarragona, NE Spain (Figure 1). All specimens were randomly collected in two sampling localities of Tarragona, distanced ~0.7 km from each other (site 1: 41°07′45″N, 1°15′28″E, 62 m a.s.l.; site 2: 41°08′6″N, 1°15′39″E; 106 m a.s.l.). Snails were gathered within a radius of ~100 m in both areas. These two sites were equivalent in terms of climate, vegetation, and soil type, and land snail assemblages present.

As expected, living specimens were only found in the fall and spring seasons, when snails from mid-latitudes are primarily active and calcification takes place. Individuals that are thought to have been aestivating or hibernating during the summer or winter, respectively, were never found despite our field sampling efforts. Collected snails were rapidly submerged in ~80% ethanol right after collection for its preservation. This sampling locality was selected because (1) relevant meteorological data are available, (2) S. candidissima living populations are abundant and accessible throughout many months of the year, and (3) Pleistocene–Holocene shell middens rich in this species are common in eastern Spain.

Body size analysis

The shell dimensions of all specimens collected in the field were measured to assess potential variations in body size throughout the year. Six standard measurements for globose shells were collected using procedures established by Kerney and Cameron (1979), including maximum shell length, width, height of the two last spires, height of the last spire, aperture height, and aperture width. All shell measurements were obtained from the field-collected 44 individuals using an electronic caliper to the nearest 0.02 mm.

Laboratory analyses

Shells were cleaned in distilled water and ultrasonication, and dried at room temperature overnight. The shell margin, which should record the environmental conditions during the last growth episode closest to snail’s collection date, was milled slowly using a manual dremel drill with a 5-mm-diameter bit. Growth lines were not visible, probably because of the continuous fast-growing mode of snail shells. Because the same researcher collected the same amount of carbonate for each shell, sampling bias are minimized. Although the time-averaging represented in the shell margin samples analyzed here is uncertain, considering that this species is short-lived and the amount of carbonate collected was very small (~150 µg), we assume that each sample represents less than ~1 month of growth.

Oxygen and carbon isotope values of the shell margin were measured using an automated carbonate preparation device (KIEL-III) coupled to a gas-ratio mass spectrometer (Finnigan MAT 252) at the University of Arizona. Powdered samples were reacted with dehydrated phosphoric acid under vacuum at 70°C. The isotope ratio measurement was calibrated using repeated measurements of NBS-19 and NBS-18. Precision was ±0.1‰ for δ¹⁸O and ±0.08‰ for δ¹³C (1-sigma). All stable isotope results are reported in δ notation relative to the international standard Pee Dee Belemnite (PDB). The δ values are defined as:

δ^{13} C or δ^{18} O = [(\frac{R_{sample}}{R_{standard}}) - 1] \times 1000 (‰)

where R = ¹³C/¹²C or ¹⁸O/¹⁶O.

Statistical analyses

All statistical analyses were performed using PAST 3.12 software (Hammer et al., 2001) considering statistical significance at α = 0.05. Linear regression analyses were carried out using the average values per collection date (n = 9). Note that snails were sometimes collected twice in a single month (a few weeks apart), which allowed us to conduct submonthly evaluations (see Table 1). A multiple regression analysis was also computed combining rain δ¹⁸O and maximum RH values.

Table 1.

Snail body size and oxygen and carbon stable isotope values of the shell margin of live-harvested adult individuals of Sphincterochila candidissima (Gastropoda: Sphincterochilidae) from Tarragona, NE Spain.

Shell ID	Date of collection	A (mm)	B (mm)	C (mm)	D (mm)	E (mm)	F (mm)	δ¹³C (PDB)	δ¹⁸O (PDB)
SC-TAR-01	8 September 2013	17.37	19.39	14.29	10.92	8.17	8.18	−9.4	0.6
SC-TAR-02	8 September 2013	16.43	20.45	13.97	10.95	8.74	8.43	−8.7	−0.4
SC-TAR-03	8 September 2013	16.45	19.75	13.40	10.47	8.28	8.15	−10.9	1.2
SC-TAR-04	8 September 2013	14.36	19.01	12.94	9.58	8.26	8.16	−9.3	1.0
SC-TAR-05	8 September 2013	13.27	17.05	11.30	8.67	6.97	7.09	−11.0	0.6
SC-TAR-06	23 October 2013	16.30	20.11	13.89	10.73	8.83	8.49	−9.6	0.2
SC-TAR-07	17 November 2013	15.13	18.94	13.24	10.08	7.93	8.03	−10.2	0.0
SC-TAR-08	17 November 2013	14.28	19.71	12.36	9.88	7.93	8.23	−10.4	−0.1
SC-TAR-09	17 November 2013	14.90	18.83	13.03	10.51	8.29	8.18	−10.8	−0.1
SC-TAR-10	30 March 2014	18.42	20.57	14.68	10.66	8.88	9.54	−5.3	0.2
SC-TAR-11	30 March 2014	18.38	20.90	15.36	10.97	9.23	9.33	−5.0	0.6
SC-TAR-12	30 March 2014	18.03	21.27	14.68	11.05	9.45	9.70	−4.9	0.5
SC-TAR-13	30 March 2014	18.45	20.91	14.60	10.83	8.40	9.11	−4.8	0.9
SC-TAR-14	30 March 2014	17.48	20.72	14.69	11.22	9.82	9.28	−5.8	1.4
SC-TAR-15	30 March 2014	17.69	21.65	15.08	11.57	8.95	9.24	−6.4	0.7
SC-TAR-16	30 March 2014	17.59	20.96	14.97	11.23	9.09	9.24	−6.3	0.8
SC-TAR-17	22 April 2014	19.84	22.16	15.90	12.01	9.89	9.39	−6.3	1.5
SC-TAR-18	22 April 2014	19.22	20.89	16.65	11.81	9.57	9.42	−5.3	1.4
SC-TAR-19	22 April 2014	18.24	19.99	15.16	11.00	9.06	9.05	−6.6	0.1
SC-TAR-20	22 April 2014	19.28	22.39	16.56	11.94	9.86	10.06	−8.1	−0.8
SC-TAR-21	22 April 2014	18.75	21.38	15.52	11.38	10.32	9.84	−8.2	1.4
SC-TAR-22	22 April 2014	19.36	21.02	15.98	11.56	9.42	9.61	−6.6	1.0
SC-TAR-23	22 April 2014	18.73	21.49	15.30	11.75	10.16	9.83	−5.8	1.0
SC-TAR-24	9 May 2014	19.47	22.37	16.19	12.24	9.98	10.25	−8.8	1.5
SC-TAR-25	9 May 2014	16.93	20.31	13.95	10.84	8.46	8.48	−10.1	0.9
SC-TAR-26	9 May 2014	19.03	18.19	14.42	9.79	8.21	8.03	−7.6	1.1
SC-TAR-27	9 May 2014	15.89	18.83	13.29	10.31	8.43	7.99	−10.4	0.0
SC-TAR-28	9 May 2014	14.89	18.88	13.52	9.76	7.52	7.68	−10.3	−0.5
SC-TAR-29	9 May 2014	15.80	17.83	12.73	9.46	7.54	7.48	−9.7	−0.5
SC-TAR-30	9 May 2014	15.89	17.87	12.89	8.19	7.54	7.55	−9.0	0.8
SC-TAR-31	15 September 2014	20.49	21.65	16.20	12.00	9.72	9.44	−6.2	1.4
SC-TAR-32	15 September 2014	18.43	20.34	15.21	11.57	9.99	8.38	−6.0	1.8
SC-TAR-33	15 September 2014	18.64	21.84	15.68	11.33	9.54	9.67	−5.4	2.3
SC-TAR-34	15 September 2014	16.94	19.96	14.14	10.57	8.76	8.43	−6.4	1.5
SC-TAR-35	28 September 2014	17.45	19.55	14.61	10.59	8.69	8.10	−6.4	3.1
SC-TAR-36	28 September 2014	16.91	20.11	14.37	10.61	7.66	7.91	−6.4	1.2
SC-TAR-37	28 September 2014	16.87	19.63	14.06	10.58	8.59	8.10	−6.1	0.7
SC-TAR-38	28 September 2014	16.39	20.61	13.47	10.36	8.20	8.51	−7.7	0.1
SC-TAR-39	28 September 2014	17.68	19.84	14.85	11.03	8.77	8.86	−8.6	1.8
SC-TAR-40	5 October 2014	18.75	22.10	15.66	12.12	9.73	9.83	−7.6	1.3
SC-TAR-41	5 October 2014	17.13	21.15	14.66	11.30	8.60	8.67	−6.6	0.7
SC-TAR-42	5 October 2014	16.77	19.37	14.49	10.25	7.92	8.54	−5.9	0.7
SC-TAR-43	5 October 2014	18.30	22.02	16.00	11.19	10.00	9.33	−6.1	−0.2
SC-TAR-44	5 October 2014	16.58	21.49	14.09	10.75	8.61	8.74	−8.6	−0.6

A: maximum shell length; B: maximum shell width; C: height of two last spires; D: height of last spire; E: aperture height; F: aperture width; PDB: Pee Dee Belemnite.

Results

Shell margin δ¹⁸O values averaged +0.7 ± 0.8‰ (n = 44) −1 sigma, and varied from +3.1‰ in September to −0.8‰ in April, so a total of 3.9‰ isotopic variability was documented between months (Table 1). Shell margin δ¹⁸O values were up to 5.7 ± 0.6‰ higher than observed local rainfall δ¹⁸O. The positive offset between shell and rain δ¹⁸O was relatively constant throughout the year. The average shell margin δ¹⁸O values per collection date (n = 9) correlated positively with monthly averaged rainwater δ¹⁸O values from the region (R² = 0.42; p = 0.05; n = 9; Figure 3a), and negatively correlated with monthly averaged maximum RH values (R² = 0.58; p = 0.018; n = 9; Figure 3b). A multiple regression analysis slightly improved the linear relationship between the dependent (shell δ¹⁸O) and these two independent (rain δ¹⁸O and max. RH) variables (R² = 0.67; p = 0.036; n = 9). No significant relationships were observed with other relevant climatic parameters for snails like average RH, air temperature, or precipitation amount. Thus, snail shell margin δ¹⁸O values exhibit a strong relationship with local maximum RH values during the snail collection date, but none with the average RH values of that month shown in Figure 2b.

Figure 3.

Relationship between shell margin δ¹⁸O values and local climate data. Filled symbols represent average values per collection date whereas whiskers represent the standard deviation of the data: (a) relationship between δ¹⁸O values of snail shell margin and precipitation; (b) relationship between δ¹⁸O values of snail shell margin and local maximum relative humidity; (c) relationship between shell margin δ¹⁸O and δ¹³C values; (d) average rainwater δ¹⁸O (open triangles) and average shell margin δ¹⁸O (filled circles) per sampling date throughout a year.

Shell margin δ¹³C values varied from −11.0‰ in September to −4.8‰ in March, so up to 6.2‰ isotopic variability was documented throughout the year (Table 1). Shell margin δ¹³C values showed an average value of −7.6 ± 1.9‰ (n = 44) and correlated positively with shell margin δ¹⁸O values (R² = 0.47; p = 0.05; n = 9; Figure 3c).

The average maximum shell length of adult S. candidissima specimens from Tarragona averaged 17.5 ± 1.6 mm (n = 44), ranging from 20.5 to 13.3 mm (Table 1). No definite body size differences were observed between seasons. However, adult snails collected in the fall of 2013 were significantly smaller (average shell length = 15.5 ± 0.8 mm) than individuals found the following fall of 2014 (average shell length = 17.7 ± 0.8 mm). Specimens from the spring of 2014 (average shell length = 18.0 ± 1.1 mm) were statistically equivalent in terms of shell dimensions than those encountered in the fall of the same year (Figure 4a). Interestingly, average shell margin δ¹⁸O and δ¹³C values for each of the nine collection dates throughout the year positively correlated with snail shell maximum length (R² = 0.41; p = 0.062; n = 9; for δ¹⁸O) and (R² = 0.78; p = 0.001; n = 9; for δ¹³C) (Figure 4b and c). Thus, larger individuals exhibited the highest δ¹⁸O and δ¹³C values in the shell margin whereas smaller specimens always showed the lowest shell margin δ¹⁸O and δ¹³C values.

Figure 4.

(a) Body size of Sphincterochila candidissima collected throughout a year in Tarragona, NE Spain. (b) Relationship between shell margin δ¹⁸O values and body size. (c) Relationship between shell margin δ¹³C values and body size. Filled circles represent average values per collection date whereas whiskers represent the standard deviation of the data.

Discussion

Shell margin oxygen isotope values

Shell margin δ¹⁸O values across seasons imitated the fluctuations in local monthly averaged rain δ¹⁸O values (Figure 3d), which reinforces that adult shells of the arid-dwelling S. candidissima from the region can be used as a convincing proxy for precipitation δ¹⁸O, even at monthly–submonthly resolution. On average, shell margin δ¹⁸O was ~5.7‰ higher than rain δ¹⁸O, an ¹⁸O-enrichment consistent with other snail species in Europe and the Mediterranean (Goodfriend et al., 1989; Lecolle, 1985). This shell–rain ¹⁸O offset was relatively constant throughout the sampled year. Balakrishnan and Yapp (2004) interpret this offset as the evaporative loss of water from the snail body fluid. Because this offset seems to be constant throughout the year (δ¹⁸O_shell–rain = 5.7 ± 0.6‰), S. candidissima shell margin δ¹⁸O values should still reflect the signature of the local rain δ¹⁸O.

The slope of the regression equation obtained in this study between rain and shell margin δ¹⁸O values of S. candidissima at the submonthly timescale (Figure 3a: δ¹⁸Op = 0.49 * δ¹⁸Os + 3.1; R² = 0.42) was comparable to other published regression equations using average annual rain and entire-shell δ¹⁸O values of different snail species combined. For instance, Zanchetta et al. (2005) measured multiple snail taxa across the Italian Peninsula (δ¹⁸Op = 0.54 * δ¹⁸Os − 5.7; R² = 0.50), Lecolle et al. (1985) analyzed numerous species around Europe (δ¹⁸Op = 1.2 * δ¹⁸Os − 5.9; R² = 0.93), and Prendergast et al. (2015) measured specimens from Northern Africa (δ¹⁸Op = 0.95 * δ¹⁸Os − 5.2; R² = 0.58). In all these case studies, including the present work here, the slope varies between ~0.5 and 1, suggesting that, on average, for every ~1‰ (SMOW) increase in rain δ¹⁸O, snail shell δ¹⁸O increases ~0.5–1‰ (PDB). The remarkable similarity in the slope of these regression equations indicates that both variables (rain and shell δ¹⁸O) exhibit similar rate of variation at both average annual and submonthly resolution. The intercept of the regression equation of this study, however, is notably different than previously published work in the Mediterranean region, suggesting that the species used here displays a significantly different offset between rain and shell δ¹⁸O than other land snail taxa. This therefore may be explained by significantly different ecologies among species and also emphasizes the need to develop site-specific and species-specific calibration studies. In contrast to other recent snail studies (Prendergast et al., 2015; Zanchetta et al., 2005), Sphincterochila δ¹⁸O values did not correlate with precipitation amount, which further reinforces the need for calibrating the target species using living individuals from the region before analyzing fossil or archeological shells for paleoclimatic inferences. Considering that the rainwater δ¹⁸O values vary seasonally, primarily in response to air temperature seasonal fluctuations, it seems possible to use the shell margin δ¹⁸O values of land snails recovered from archeological sites to estimate, at least roughly, the approximate season of snail collection. However, these kinds of studies appear to be more plausible in lower-latitude areas where snails may grow continuously year-round and shell margin δ¹⁸O values track rain δ¹⁸O values across all seasons, rather that just fall and spring seasons as documented here. This hypothesis remains to be tested in future studies in subtropical–tropical areas.

Both land snails and slugs are highly sensitive to dehydration via water loss through evaporation under dry conditions. While in moist environments, temperature seems to be the limiting factor for snails, in dry environments, moisture is the major controlling factor for snail active periods (Moreno-Rueda, 2007; Moreno-Rueda and Collantes-Martin, 2007). Hence, it is expected that snail active periods and therefore most of shell growth, would occur at night, when RH is the highest, especially in arid environments. The snail flux balance model of Balakrishnan and Yapp (2004), combined with the aragonite–water oxygen isotope fractionation equation of Grossman and Ku (1986), was used to infer RH values from the shell margin δ¹⁸O values (Figure 5). The model posits that temperature, δ¹⁸O values of water and water vapor, and RH are the most important variables in the accurate determination of the body fluid and the shell δ¹⁸O value (Balakrishnan and Yapp, 2004). Another important parameter is the flux of liquid water output from the body fluid (f_o) relative to the flux of liquid water ingested by the snail (f_in) during shell formation. This ratio is called θ, which is defined as θ = (f_o/f_in). Balakrishnan and Yapp (2004) showed that it is appropriate to assume that ambient water vapor is in isotope equilibrium with the liquid water. Furthermore, for θ < 0.40, an assumption that snail body fluid is lost only by evaporation (i.e. θ = 0) is a good approximation and introduces very little error in model calculations (Balakrishnan and Yapp, 2004). Thus, a value of θ = 0 is adopted here.

Figure 5.

(a) Calculations of the shell margin δ¹⁸O values as a function of RH using the snail evaporative steady-state flux balance model by Balakrishnan and Yapp (2004). Filled circle depicts the average shell margin δ¹⁸O value of all 44 analyzed snails in this study. (b) Relationship between calculated RH from the snail shell margin δ¹⁸O values and observed maximum RH values at the snail collection date.

The average shell margin δ¹⁸O value of Sphincterochila individuals collected in the fall and spring seasons in Tarragona showed an average value of +0.7 ± 0.8‰ versus PDB (n = 44). If collected snails grew their shell lip at average atmospheric conditions for the fall and spring seasons, that is, air temperature of ~16°C and rainwater δ¹⁸O values of ~ −5‰ (SMOW), then, the model predicts that snails were active at times when RH was, on average, ~89%. The predicted RH value from the snail model is a value very close to the observed average maximum RH value (~90%) for those same months combined. Accordingly, the shell δ¹⁸O values of S. candidissima shells seem to credibly record maximum RH values in the context of the flux balance model by Balakrishnan and Yapp (2004), which is constrained by independent climate data.

Model calculations were also used to predict RH values throughout the sampled months. The calculated RH values per collection date, assuming that calcification occurred under atmospheric conditions operating during the same collection date, positively correlated with observed maximum RH values (R² = 0.61; p = 0.016; n = 9). The significant relationship between calculated and observed RH values through seasons further reinforces the usefulness of the model by Balakrishnan and Yapp (2004) to predict moisture conditions during snail calcification when other climatic variables (shell δ¹⁸O, rainwater δ¹⁸O, and air temperature) are known or can be deduced independently. Overall, the shell δ¹⁸O values of S. candidissima appear to respond primarily to variations in rain δ¹⁸O and maximum RH values, with higher shell δ¹⁸O values associated with somewhat drier conditions (i.e. higher rain δ¹⁸O values and/or lower RH) while lower values are linked to wetter times (i.e. lower rain δ¹⁸O values and/or higher RH). Future studies in the region will evaluate the validity of shell margin δ¹⁸O values from Pleistocene–Holocene shell middens to attempt to reconstruct atmospheric conditions at the time when snails were harvested by ancient human groups.

Shell margin carbon isotope values

The δ¹³C values of land snail shells are primarily a function of the consumed and assimilated plants, as shown in cultured experiments (Metref et al., 2003; Stott, 2002; Zhang et al., 2014). Considering that there is an isotopic offset between shell and diet of ~17‰ for this species (Yanes et al., 2013a), and that the average shell margin δ¹³C value for Tarragona specimens is ~ −7.6‰, sampled snails should have ingested (at the time of collection) resources with an average δ¹³C value of ~ −24.6‰, ranging from ~ −21.8‰ for snails with a shell δ¹³C value of −4.8‰ to ~ −28.0‰ for individuals with a shell δ¹³C value of −11.0‰. Accordingly, we can conclude that measured specimens followed a pure C₃ plant diet throughout the year, which is reasonable considering that the sampling area is dominated by Mediterranean C₃-type plants.

Even though it has been shown that many small land snail taxa do not seem to incorporate significant amounts of limestone in the shell (Pigati et al., 2004, 2010), for many large body size species with access to limestone, the ingestion of foreign carbonates can be significant and detectable in the shell via radiocarbon and stable carbon isotope analysis (Goodfriend et al., 1999; Yanes et al., 2008). S. candidissima is known to be abundant in soils and areas rich in calcium carbonate, and considering the large size and thickness of its white shell, ingestion of carbonate as a source of calcium for shell growth is probably important (Yanes et al., 2013a), possibly accounting for the ~10–30% of snail shell carbon isotope budget (Goodfriend et al., 1999; Yanes et al., 2008). Although the shell δ¹³C values are overwhelmingly dominated by the δ¹³C values of consumed plants, the effect of limestone ingestion is possibly also recorded in the shell of Sphincterochila as suggested by carbon flux balance mixing models for the same species in other localities in Spain (see Figure 3 in Yanes et al., 2013a). However, because limestone δ¹³C values were not available for the present study, further quantitative evaluations are not yet possible. Accordingly, higher δ¹³C values in S. candidissima shells should reflect C₃ plants enriched in ¹³C, possibly at times of drier conditions when plants are more water-stressed (Dawson et al., 2002; Farquhar et al., 1989; Prendergast et al., in press), perhaps in combination with increased limestone intake. In contrast, lower shell δ¹³C values are related to C₃ plants with lower δ¹³C values, likely in response to wetter conditions, and/or lower limestone intake. Even though some studies have speculated that shell δ¹³C values can also be impacted by atmospheric CO₂ (e.g. ZongXiu et al., 2007), a published flux balance mixing model suggests that the δ¹³C values of well-mixed atmospheric CO₂ should not significantly impact the snail carbon isotope budget because of the small difference in the partial pressure of the CO₂ in the lungs of the snail and the open atmosphere (Balakrishnan and Yapp, 2004).

Considering that oxygen and carbon stable isotopes of S. candidissima positively correlate (Figure 3c), which is unusual in other edible pulmonate snail species (see Figure 6 in Yanes et al., 2013a), both isotope systems appear to be directly or indirectly affected by moisture. Thus, Sphincterochila shell δ¹³C values may be used as an additional indirect proxy for moisture, although this might not be the case for most land snail species.

Relationship between stable isotopes and body size

In the present work, we observed a positive correlation between Sphincterochila shell size and both oxygen and carbon stable isotopes (Figure 4), which suggests that body size of this species may respond to both climate and diet. It is generally anticipated that snail body size is significantly influenced by climate. However, the extent and direction to which climate influences snail ultimate body size is unclear. In arid regions, moisture is the dominant limiting factor for land snails. While some studies have observed that snails reach larger sizes in desert areas as a survival mechanism to minimize the risk of water loss because it decreases the surface-to-volume ratio (Nevo et al., 1983), other studies have observed that snail size tends to increase with increasing moisture because snails are able to grow and be active for longer periods (Goodfriend, 1986). A third possibility is that snail size does not exhibit a linear relationship with rainfall (Hausdorf, 2006).

Considering that shell δ¹⁸O values of Sphincterochila primarily reflect moisture conditions, with higher values mainly associated with higher rain δ¹⁸O values and lower RH (Figure 3), we can infer that S. candidissima individuals reach larger size during drier times. Therefore, Sphincterochila ultimate body size may function as an ecological mechanism to survive drier conditions by lowering the surface-to-volume ratio and thus, reduce the risk of water loss by evaporation. The same observations were made by Yanes et al. (2012) for the edible land snail Cepaea nemoralis (Gastropoda: Helicidae) from coastal sites in Asturias and Cantabria, northern Spain.

The positive relationship between snail body size and shell δ¹³C values (Figure 4) also suggest that adult individuals that ingested more water-stressed plants, which are usually enriched in ¹³C (Dawson et al., 2002; Farquhar et al., 1989), and/or higher proportion of limestone, which displays significantly higher δ¹³C values than C₃ plants (Yanes et al., 2013a), tend to reach larger dimensions in the study area.

All in all, the results of this study support the hypothesis that adult specimens of S. candidissima from Spain tend to exhibit larger dimensions under drier conditions, possibly as a survival mechanism to reduce the surface-to-volume ratio. Correspondingly, shell size of adult individuals may be used as an additional proxy for moisture conditions in the western Mediterranean. Future studies in the region will analyze body size and isotope composition of the shell jointly using archeological specimens to further evaluate this hypothesis using ancient shelly materials. Although this approach does not seem to be valid for all snail species (Hausdorf, 2006), the shell size of S. candidissima and other edible species like the European snail Cepaea nemoralis may be used as a complementary proxy for humidity, with larger individuals pointing to somewhat drier conditions.

Archaeological implications

Although earlier archaeological work generally focused on zooarcheological remains (e.g. bones, marine shells) other than land snails, recent investigations illustrate that (1) land snails have been consistently exploited by many prehistoric foragers (Bosch et al., 2015; Fernández-Lopez-de-Pablo et al., 2014; Hill et al., 2015; Lubell, 2004a, 2004b); and that (2) the stable isotope composition of archeological shells can credibly reflect average (dominant) climatic conditions during relevant cultural periods (e.g. Colonese et al., 2007, 2010a, 2010b, 2011, 2013a, 2013b; Paul and Mauldin, 2013; Prendergast et al., 2016a; Stevens et al., 2012; Yanes et al., 2011, 2012, 2013a, 2013b, 2014). Studies that assess the season of shell harvest are also relevant for archeologists because they inform about season of site occupation, human mobility, and subsistence patterns (e.g. Prendergast et al., 2016b). Our results demonstrate that shell margin δ¹⁸O values of S. candidissima from northeastern Spain (~41° N) convincingly record the dominant conditions at the time of snail collection, that is, at monthly–submonthly resolution (Figure 3a, b and d). However, because the target species seems to only grow during fall and spring seasons (only half of the months of the year), estimates of season of collection during prehistoric times seem limited in this case study.

Another difficulty is whether or not snails harvested by human foragers were collected while specimens were active, hibernating/aestivating, or both. If snails were collected at times when snails did not secrete shell, then, inferences of season of shell collection may be biased. Since S. candidissima remains inactive during the winter and summer seasons, and consequently, does not grow shell, it could be argued that the archaeological interest of this species to determine the season of collection is of limited use. Our current knowledge about cultural exploitation and density of S. candidissima suggests that humans likely harvested snails during snail active periods (fall and spring). Ethno-biological references about the human collection of land snails in Mediterranean Spain indicates that the collection takes place after rain events, usually in April and May, when snails are most active (Fajardo et al., 2009). In addition, field observations in Mediterranean Spain suggest that S. candidissima displays significantly higher abundance after rain events, and therefore, they are easier to find and collect while they are active rather than when they are aestivating/hibernating (personal field observations, 2013–2014). We therefore assume that ancient human groups likely harvested snails more easily while they were active.

The fact that shell margin δ¹⁸O values reflect dominant conditions at the time of snail collection, even for fall and spring seasons only, is a very encouraging result and calls for additional investigation. This sampling approach to determine season of shell harvest may be more effective in studies from lower-latitude regions where snails may grow continuously and uninterruptedly year-round.

This study also illustrates that body size of S. candidissima snails can be used as a qualitative proxy for moisture in the region, given the documented associations between shell margin isotope composition and shell dimensions (Figure 4b and c). This observation could be explored and applied to archeological studies that use the same species. However, if past environments were significantly different than at present-day, body size of archeological shells may differ from those of today (Yanes et al., 2012). In those cases, the combination of body size and stable isotopes from the same shells will be critical to determine if these relationships hold in the past.

Conclusion

This is the first study to investigate the isotopic composition of the last growth episode (shell lip) of adult land snails and its relationship to climate throughout a year. Arid-dwelling S. candidissima was collected during 2013–2014 in Tarragona, NE Spain. The results show that shell margin δ¹⁸O values primarily record local rainwater δ¹⁸O values and maximum RH conditions, reinforcing that S. candidissima can certainly be used as a valid paleoprecipitation δ¹⁸O and/or paleohumidity proxy at both annual and monthly–submonthly resolution. The positive relationship between shell δ¹⁸O, δ¹³C, and shell size suggests that specimens of Sphincterochila tend to reach larger size under drier conditions, probably as a mechanism to reduce the risk of water loss associated with lower surface-to-volume ratio. The size of this species, hence, may be used as a complementary paleohumidity proxy. The shell margin δ¹³C values varied significantly throughout the year suggesting significant variations in snail diet and limestone ingestion. The observed strong correlation between shell δ¹³C and δ¹⁸O values, not common for other edible snail species, suggests that both isotope systems are influenced by moisture directly or indirectly.

Footnotes

Acknowledgements

Special thanks go to the detailed and thoughtful comments provided by three anonymous reviewers who critically evaluated this work.

Funding

This research has been supported by the Spanish Ministerio de Economía y Competitividad (MICINN) grants RYC-2011-09363, HAR2013-41197-P to JFL and the National Science Foundation (NSF) grant EAR-1529133 to YY.

References

Andrus

CFT

(2011) Shell midden sclerochronology. Quaternary Science Reviews 30: 2892–2905.

Andrus

CFT

Thompson

(2012) Determining the habitats of mollusk collection at the Sapelo Island shell ring complex, Georgia, USA using oxygen isotope sclerochronology. Journal of Archaeological Science 39: 215–228.

Aparicio

Álvarez-Fernández

Jordà-Pardo

. (2015) Análisis de los moluscos continentales del Paleolítico Superior y del Epipaleolítico de la Cueva de Nerja (sala de la Mina), Málaga, España. In: Gutiérrez-Zugasti

Cuenca-Solana

González-Morales

(eds) La Investigación Arqueomalacológica en la Península Ibérica: Nuevas Aportaciones. Santander: Nadir Ediciones, pp. 27–40.

Araguas-Araguas

Diaz-Teijeiro

(2005) Isotopic composition of precipitation in the Mediterranean basin in relation to air circulation patterns and climate. IAEA-TECDOC-1453, Vienna, pp. 173–190. Available at: http://www-pub.iaea.org/MTCD/publications/PDF/te_1453_web.pdf.

Balakrishnan

Yapp

(2004) Flux balance model for the oxygen and carbon isotope compositions of land snail shells. Geochimica et Cosmochimica Acta 68: 2007–2024.

Balakrishnan

Yapp

Meltzer

. (2005) Paleoenvironment of the Folsom archaeological site, New Mexico, USA, approximately 10,500 14C yr B.P. as inferred from the stable isotope composition of fossil land snail shells. Quaternary Research 63: 31–44.

Bosch

Wessenlingh

Mannino

(2015) The Ksâr ’Akil (Lebanon) mollusc assemblage: Zooarchaeological and taphonomic investigations. Quaternary International 390: 85–101.

Colonese

Troelstra

Ziveri

. (2009) Mesolithic shellfish exploitation in SW Italy: Seasonal evidence from the oxygen isotopic composition of Osilinus turbinatus shells. Journal of Archaeological Science 36(9): 1935–1944.

Colonese

Zanchetta

Dotsika

. (2010a) Early–middle Holocene land snail shell stable isotope record from Grotta di Latronico 3 (southern Italy). Journal of Quaternary Science 25: 1347–1359.

10.

Colonese

Zanchetta

Fallick

. (2007) Stable isotope composition of late glacial land snail shells from Grotta del Romito (Southern Italy): Paleoclimatic implications. Palaeogeography, Palaeoclimatology, Palaeoecology 254: 550–560.

11.

Colonese

Zanchetta

Fallick

. (2010b) Stable isotope composition of Helix ligata (Muller, 1774) from Late Pleistocene–Holocene archaeological record from Grotta della Serratura (Southern Italy): Paleoclimatic implications. Global and Planetary Change 71: 249–257.

12.

Colonese

Zanchetta

Fallick

. (2013a) Holocene snail shell isotopic record of millennial-scale hydrological conditions in western Mediterranean: Data from Bauma del Serrat del Pont (NE Iberian Peninsula). Quaternary International 303: 43–53.

13.

Colonese

Zanchetta

Perlès

. (2013b) Deciphering late Quaternary land snail shell d¹⁸O and d¹³C from Franchthi Cave (Argolid, Greece). Quaternary Research 80: 66–75.

14.

Colonese

Zanchetta

Russell

. (2011) Stable isotope composition of Late Pleistocene–Holocene Eobania vermiculata shells (Müller, 1774) (Pulmonata, Stylommatophora) from the Central Mediterranean basin: Data from Grotta d’Oriente (Favignana, Sicily). Quaternary International 244: 76–87.

15.

Dawson

Mambelli

Plamboeck

. (2002) Stable isotopes in 568 plant ecology. Annual Review of Ecology and Systematics 33: 507–559.

16.

Fajardo

Verde

Obón

. (2009) La recolección y consumo tradicional de caracoles terrestres en Albacete. Sabuco 7: 235–264.

17.

Farquhar

Ehleringer

Hubick

(1989) Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40: 503–537.

18.

Fernández-Lopez-de-Pablo

Gabriel

(2016) El Collado shellmidden and the exploitation patterns of littoral resources during the Mesolithic in the Eastern Iberian Peninsula. Quaternary International 407B: 106–117.

19.

Fernández-Lopez-de-Pablo

Badal

Ferrer-García

. (2014) Land snails as a diet diversificaion proxy during the early upper palaeolithic in Europe. PLoS ONE 9(8): e104898.

20.

Fernández-Lopez-de-Pablo

Gómez-Puche

Martınez-Ortí

(2011) Systematic consumption of non-marine gastropods at open air Mesolithic sites in the Iberian Mediterranean region. Quaternary International 244: 45–53.

21.

Goodfriend

(1986) Variation in land snail shell form and size and its causes: A review. Systematic Zoology 35: 204–223.

22.

Goodfriend

Ellis

Toolin

(1999) Radiocarbon age anomalies in land snail shells from Texas: Ontogenetic, individual and geographic patterns of variation. Radiocarbon 41: 149–156.

23.

Goodfriend

Magaritz

Gat

(1989) Stable isotope composition of land snail body water and its relation to environmental waters and shell carbonate. Geochimica et Cosmochimica Acta 53: 3215–3221.

24.

Grossman

(1986) Oxygen and carbon isotope fractionation in biogenic aragonite temperature effects. Chemical Geology 59: 59–74.

25.

Hallmann

Burchell

Schöne

. (2009) High resolution sclerochronological analysis of the bivalve mollusk Saxidomus gigantea from Alaska and British Columbia: Techniques for revealing environmental archives and archaeological seasonality. Journal of Archaeological Science 36: 2353–2364.

26.

Hammer

Harper

DAT

Ryan

(2001) PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4(1): 9.

27.

Hausdorf

(2006) Is the interspecific variation of body size of land snails correlated with rainfall in Israel and Palestine? Acta Oecol 30: 374–379.

28.

Hill

Hunt

Lucarini

. (2015) Land gastropod piercing during the Late Pleistocene and Early Holocene in Hauah Fteah, Lybia. Journal of Archaeological Science 4: 320–325.

29.

Hutterer

Linstädter

Eiwagner

. (2014) Human manipulation of terrestrial gastropods in Neolithic culture groups of Morocco. Quaternary International 320: 83–91.

30.

Kerney

Cameron

RAD

(1979) A Field Guide to the Land Snails of Britain and North-West Europe. London: William Collins, Sons.

31.

Lecolle

(1985) The oxygen isotope composition of land snail shells as a climatic indictor: Applications to hydrogeology and palaeoclimatology. Chemical Geology 58: 157–181.

32.

Liu

. (2007) Diet control on carbon isotopic composition of land snail shell carbonate. Chinese Science Bulletin 52: 388–394.

33.

Lubell

(2004a) Are land snails a signature for the Mesolithic–Neolithic transition? Documenta Praehistorica 31: 1–24.

34.

Lubell

(2004b) Prehistoric edible land-snails in the circum-Mediterranean: The archaeological evidence. In: Brugal

Desse

(eds) Petits Animaux et Societes Humaines. Du Complement Alimentaire aux Ressources Utilitaires. Antibes: Editions APDCA, pp. 77–98.

35.

Mannino

Mazzanti

(2013) Studi ed osservazioni preliminary sulla malacofauna dal sitio capsiano costiero SHM-1 (Tunisia centrale). In: Mulazzani

(ed.) Le Capsien de Hergla (Tunisie): Culture, environnement et économie: Projet italo-tunisien de recherches archéologiques sous la direction de R. Boussoffara, S. Mulazzani et M. Tosi. Frankfurt am Main: Africa Magna Verlag, pp. 339–361.

36.

Mannino

Thomas

Leng

. (2007) Marine resources in the Mesolithic and Neolithic at the Grotta dell’Uzzo (Sicily): Evidence from isotope analyses of marine shells. Archaeometry 49: 117–133.

37.

Metref

Rousseau

Bentaleb

. (2003) Study of the diet effect on d¹³C of shell carbonate of the land snail Helix aspersa in experimental conditions. Earth and Planetary Sciences Letters 211: 381–393.

38.

Moreno-Rueda

(2007) Refuge selection by two sympatric species of arid dwelling land snails: Different adaptive strategies to achieve the same objective. Journal of Arid Environments 68: 588–598.

39.

Moreno-Rueda

(2008) The white colour diminishes the weight loss during aestivation in the arid-dwelling land snail Sphincterochila (Albea) candidissima. Iberus 26: 47–51.

40.

Moreno-Rueda

(2014) Distribution of arid-dwelling land snails according to dryness. Journal of Arid Environments 103: 80–84.

41.

Moreno-Rueda

Collantes-Martin

(2007) Annual cycle of activity of the land-snail Sphincterochila (Albea) candidissima (Draparnaud, 1801) in a semi-arid environment. Iberus 25: 49–56.

42.

Nevo

Bar-El

Bar

(1983) Genetic diversity, climatic selection and speciation of Sphincterochila land snails in Israel. Biological Journal of the Linnean Society 19: 339–373.

43.

Paul

Mauldin

(2013) Implications for Late-Holocene climate from stable carbon and oxygen isotopic variability in soil and land snail shells from archaeological site 41KM6 9 in Texas, USA. Quaternary International 308–309: 242–252.

44.

Pigati

Quade

Shanahan

. (2004) Radiocarbon dating of minute gastropods and new constraints on the timing of spring-discharge deposits in southern Arizona, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 204(1–2): 33–45.

45.

Pigati

Rech

Nekola

(2010) Radiocarbon dating of small terrestrial gastropod shells in North America. Quaternary Geochronology 5: 519–532.

46.

Prendergast

Azzopardi

O’Connell

. (2013) Oxygen isotopes from Phorcus (Osilinus) turbinatus shells as a proxy for sea surface temperature in the central Mediterranean: A case study from Malta. Chemical Geology 345: 77–86.

47.

Prendergast

Stevens

Barker

. (2015) Oxygen isotope signatures from land snail (Helix melanostoma) shells and body fluid: Proxies for reconstructing Mediterranean and North African rainfall. Chemical Geology 409: 87–98.

48.

Prendergast

Stevens

Hill

. (in press) Carbon isotope signatures from land snail shells: Implications for palaeovegetation reconstruction in the eastern Mediterranean. Quaternary International. DOI: 10.1016/j.quaint.2014.12.053.

49.

Prendergast

Stevens

O’Connell

. (2016a) A late Pleistocene refugium in Mediterranean North Africa? Palaeoenvironmental reconstruction from stable isotope analysis of land snail shells (Haua Fteah, Libya). Quaternary Science Reviews 139: 94–109.

50.

Prendergast

Stevens

O’Connell

. (2016b) Changing patterns of eastern Mediterranean shellfish exploitation in the Late Glacial and Early Holocene: Oxygen isotope evidence from Phorcus turbinatus shells in Epipaleolithic to Neolithic human occupation layers at the Haua Fteah cave, Libya. Quaternary International 407: 80–93.

51.

Stevens

Metcalfe

Leng

. (2012) Reconstruction of Late Pleistocene climate in the Valsequillo Basin (Central Mexico) through isotopic analysis of terrestrial and freshwater snails. Palaeogeography, Palaeoclimatology, Palaeoecology 319–320: 16–27.

52.

Stott

(2002) The influence of diet on the d¹³C of shell carbon in the pulmonate snail Helix aspersa. Earth and Planetary Sciences Letters 195: 249–259.

53.

Surge

Barrett

(2012) Marine climatic seasonality during medieval times (10th to 12th centuries) based on isotopic records in Viking Age shells from Orkney, Scotland. Palaeogeography, Palaeoclimatology, Palaeoecology 350–352: 236–246.

54.

Yanes

Delgado

Castillo

. (2008) Stable isotope (δ¹⁸O, δ¹³C, and δD) signatures of recent terrestrial communities from a low-latitude, oceanic setting: Endemic land snails, plants, rain, and carbonate sediments from the eastern Canary Islands. Chemical Geology 249: 277–292.

55.

Yanes

Gómez-Puche

Esquembre-Bebia

. (2013a) Younger Dryas – Early Holocene transition in southeastern Iberian Peninsula: Insights from land snail shell middens. Journal of Quaternary Science 28: 777–788.

56.

Yanes

Gutierrez-Zugasti

Delgado

(2012) Late glacial–early Holocene transition in northern Spain deduced from terrestrial gastropod shelly accumulations. Quaternary Research 78: 373–385.

57.

Yanes

Izeta

Cattaneo

. (2014) Holocene (~4.5–1.7 cal. Kyr BP) paleoenvironmental conditions in central Argentina inferred from entire-shell and intra-shell stable isotope composition of terrestrial gastropods. The Holocene 24: 1193–1205.

58.

Yanes

Riquelme

Camara

. (2013b) Stable isotope composition of middle to late-Holocene land snail shells from the Marroquíes archeological site (Jaen, Southern Spain): Paleoenvironmental implications. Quaternary International 302: 77–87.

59.

Yanes

Romanek

Molina

. (2011) Holocene paleoenvironment (7,200–4,000 cal. years BP) of the Los Castillejos archaeological site (southern Spain) as inferred from stable isotopes of land snail shells. Quaternary International 244: 67–75.

60.

Zanchetta

Leone

Fallick

. (2005) Oxygen isotope composition of living land snail shells: Data from Italy. Palaeogeography Palaeoclimatology Palaeoecology 223: 20–33.

61.

Zhang

Yamada

Suzuki

. (2014) Factors controlling shell carbon isotopic composition of land snail Acusta despecta sieboldiana estimated from lab culturing experiment. Biogeosciences 11: 6555–6590.

62.

ZongXiu

ZhaoYan

NaiQin

. (2007) Diet control on carbon isotopic composition of land snail shell carbonate. Chinese Science Bulletin 52: 388–394.