Introduction

Sells of freshwater bivalve Dreissena polymorpha (Pallas) are an important constituent of the topmost deposits of numerous lakes. Empty shells may constitute more than 90 % of the deposits within the sublittoral (Lewandowski 2001). Due to its wide spatial range, rapid dispersal, massive occurrence and resistance to water pollution, D. polymorpha may be a good source of information about environmental conditions over vast areas.

The species has already gained much attention from a broad range of scientists, including biologists, ecologists and limnologists (van der Velde et al. 2010). However, despite its invasive character and its ability to outcompete native species in an increasing number of areas, D. polymorpha shells have not received much attention in isotope studies. In a detailed study, Wurster and Patterson (2001) found δ18O in D. polymorpha shells to be in equilibrium with δ18Owater, showing that the stable oxygen isotope record can serve as a palaeoclimate proxy. Fry and Allen (2003) established that the stable carbon (C), nitrogen (N) and sulphur (S) isotope compositions of zebra mussel inhabiting the Mississippi River were changing seasonally and were dependent on the shifts in watershed inputs and chemistry of the river water.

The interest in D. polymorpha, commonly known as the zebra mussel, has accelerated since the species reached North America in 1988 and quickly spread into the northeastern part of the United States (Mackie and Schlosser 1996). D. polymorpha is now listed among 100 of the World’s Worst Invasive Alien Species (Global Invasive Species Database 2012).

Originally native to the Black, Caspian and Azov Seas, D. polymorpha is now a widely distributed species (Stańczykowska et al. 2010). Pallas (1771) was the first to describe the species from the Caspian Sea and the Ural River. At the turn of the 18th and 19th centuries, zebra mussels had spread to most major drainages of Europe because of the widespread construction of canal systems (Bij de Vaate et al. 2002). Since then, the species has expanded its range to most of Europe. Its rapid dispersal throughout major rivers connected with canals was due to passive drifting of the larval stage (the free-floating or ‘pelagic’ veliger) and its ability to attach to boats (USGS 2012). The spread of D. polymorpha continues. The release of larval mussels during the ballast exchange of a single commercial cargo ship can introduce the species to new areas. This spread is enhanced because under cool, humid conditions, zebra mussels can stay alive for several days out of water (USGS 2012). Due to its ease in adapting to new habitats and due to the high number of individuals, D. polymorpha is displacing native species (Birnbaum 2011), causing problems for the biodiversity of natural habitats.

In the past two decades, there have been an increasing number of papers describing the relationship of δ13C and δ18O with mollusc shells and ambient waters (Fastovsky et al. 1993; Dettman et al. 1999; Kaandorp et al. 2003; Yoshimura et al. 2010; Schöll-Barna et al. 2012). The main aim of these studies was to answer the question of whether stable carbon and oxygen isotope compositions recorded in the shells of different freshwater mollusc species may be successfully used in palaeolimnological research.

However, another question arises: Do coeval shells within one population have equal isotope signatures? In the present paper, the isotope values of individual D. polymorpha shells from a population in Lake Lednica, western Poland (Fig. 1), are described. The main aim of the present study was to determine the range of the carbon and oxygen isotope values in coeval shells from one population. The isotope composition of the shells is also compared with the isotope composition of the waters.

Fig. 1
figure 1

A Map showing the location of Lake Lednica in Poland. B Detailed map of Lake Lednica with sampling sites (L6, L7, L8, L9) indicated with black dots

Full size image

Study site

Lake Lednica is situated in the southern part of the Gniezno Lake District, approximately 35 km east of the city of Poznań in west-central Poland (52°33′N, 17°23′E; Fig. 1). It is an elongated lake with an area of ~3.4 km2, and it fills the southern part of a tunnel valley extending between Janowiec and Lednogóra. The basin is ~7300 m long and has a maximum width of 825 m and a maximum depth of 15.1 m. Apart from direct precipitation, limited surface run-off and groundwater input, the lake is fed by several small, temporary streams that flow mainly during the early spring and after events of intense precipitation and by one permanent surface inlet in the southeastern part of the lake (Kolendowicz 1992). The lake catchment is relatively small, at ~38 km2. Before the 1990s, the lake was hydrologically open, with an outflow through the small River Główna in the southern part of the lake; however, this connection no longer exists. Due to the insignificant inflow compared to the lake’s volume of ~2.4 × 107 m3, the modern water residence time of the lake is relatively long and has been estimated as approximately 6 years (Jańczak 1991), in agreement with an estimated yearly water exchange of 18 % (Tybiszewska and Szulczyńska 2003). The relief of the study area was formed during the Weichselian glaciation. Aspects of the glacial and postglacial history of the area were summarised in Apolinarska and Ciszewska (2006).

The study area is influenced by both the Atlantic and continental air masses, with the former commonly prevailing. The present climate is characterised by a mean annual precipitation of ~500 mm and a mean annual temperature of 7.9 °C (mean July 17.9 °C; mean January −2.4 °C). A seasonal precipitation minimum causes an effective moisture deficit during the summer months (Kondracki 2000).

Materials and methods

Macrophytes occurring within the littoral of Lake Lednica with D. polymorpha individuals attached were collected using an Ekman bottom grab sampler (AMS 445.11; dimensions 0.15 × 0.15 × 0.15 m) in June and August 2011. The samples were placed in plastic bags and transported to the Institute of Geology, Poznań.

At each sampling site, water samples for stable carbon and oxygen isotope analyses were collected into two 10-ml glass septa test tubes at a depth of 0.5 m and directly above the macrophytes using a bathometer (5-liter Uwitec Plexiglass Watersampler). The water samples were preserved with two drops of HgCl2. Dissolved oxygen concentrations and water temperatures were measured at the same sites and depths using an Elmetron CX-401 portable meter, and electrolytic conductivity was measured using a Cyber-Scan 200.

In June, the samples were taken from four transects established within the lake’s littoral zone (L6, L7, L8, L9); in August, only two transects were selected (L6, L7), according to D. polymorpha distribution observed in June (Fig. 1; Table 1).

Table 1 Number of D. polymorpha individuals occurring at macrophytes calculated per square meter (m2)
Full size table

In the laboratory, all live D. polymorpha individuals present on macrophytes were handpicked. The molluscs were treated with about 50 % alcohol for conservation until further treatment. Soft parts were handpicked with a pair of tweezers. Empty shells were placed in 10 % H2O2 for 48 h to eliminate organics that might interfere with the isotope results. Eventually, 210 D. polymorpha shells were homogenised individually in an agate mortar and placed in Eppendorf vials.

The stable carbon and oxygen isotope compositions of the mollusc shells and water samples were analysed at the Isotope Dating and Environment Research Laboratory in Warsaw, Poland. Carbonates were dissolved with 100 % phosphoric acid (density 1.9) at 75 °C using a Kiel IV online carbonate preparation line connected to a ThermoFinnigan Delta + mass spectrometer. All values are reported as δ values, where δ = (Rsample/Rstandard − 1) × 1,000, in per mil relative to V-PDB by assigning a δ13C value of 1.95 ‰ and a δ18O value of −2.20 ‰ to NBS19. Reproducibility was checked on the basis of the long-term repeatability of NBS19 analysis and was better than ± 0.07 and 0.12 ‰ for δ13C and δ18O, respectively.

Isotope analyses of both dissolved inorganic carbon (δ13CDIC) and water (δ18Owater) were conducted using a GasBench-II headspace autosampler online with a Finnigan MAT 253 isotope ratio mass spectrometer (IRMS). During the δ13CDIC determination procedure septum-sealed sample vials were first filled with 3–4 drops of phosphoric acid (98 %) and then flushed with a continuous flow of He. Then the samples were injected into the vials with a syringe and CO2 was released due to contact with acid. The CO2 and He mixture was left to equilibrate for 18 h and after purification (removal of water vapour) and separation on a GC column, CO2 was measured in the IRMS using sample/reference gas comparison. To ensure the precision of the results four international carbonate standards were measured in each series of samples: NBS 18, NBS 19, LSVEC and IAEA-CO-9.

For the oxygen isotope analyses the vials were filled with 0.5 ml of samples with disposable pipettes and sealed with septa. The air was then removed from the vials by an automated, autosampler-assisted procedure which uses a gas mixture of CO2 and He. CO2 serves as an equilibration gas. After the required equilibration time (18 h) between the oxygen from the sample and from CO2, the gas was purified (of water vapour), separated on a GC column and measured on the IRMS. To ensure the precision of the results, three international standards (VSMOW, GISP, SLAP) and one internal standard were measured in each series of samples.

Results

Basic physico-chemical parameters of the water

The temperature of the surface waters in June ranged between 21.3 and 23 °C, however, values exceeding 22.6 °C were restricted to an isolated bay in the southern part of lake, which has limited wind action (Transect L9, Fig. 1; Table 1). In August, the surface water temperatures ranged between 21.1 and 22.1 °C (Transect L9 was excluded from the field investigations, Table 1).

In June, the difference in water temperature between the surface and bottom samples (up to 5 m of depth) did not exceed 0.7 °C; in most samples, however, it ranged between 0.3 and 0.5 °C. A significant decrease in the bottom water temperature was observed in August. In the samples derived from depths of 6 and 7 m, the water temperature was 20.1 and 18.6 °C, respectively, which is lower than the temperature in the surface waters by 1.3 and 3.1 °C, respectively.

The dissolved oxygen concentrations varied within a significant range of values, i.e., between 5.9 and 13.7 mg l−1. The highest concentrations were observed in the surface waters within the shallow littoral. The electrolytic conductivity was between 733 and 773 μS cm−1 in June and 777 and 810 μS cm−1 in August.

Stable isotope composition of Lake Lednica waters

The stable carbon isotope values in the dissolved inorganic carbon (DIC) of the lake surface and bottom waters ranged between −6.77 and −5.99 ‰ and −6.72 and −5.62 ‰, respectively, in the June samples and between −5.99 and −5.58 ‰ and −8.7 and −5.61 ‰, respectively, in the August samples (Fig. 2). The diversity of δ18O in the lake surface and bottom waters is insignificant in the June samples, i.e., −4.50 to −4.35 ‰ and −4.49 to −4.33 ‰, respectively, and in the August samples, i.e., −4.25 to −4.15 ‰ and −4.28 to −3.97, respectively.

Fig. 2
figure 2

Cross plot of δ13CDIC and δ18Owater values in Lake Lednica. Empty symbols (open circle, open square, open triangle, open diamond)—surface water samples. Filled symbols (filled circle, filled square, filled triangle, filled diamond)—bottom water samples. Dashed line separates isotope results of June and August water samples

Full size image

During the vegetation season, the surface lake waters evolved towards being isotopically heavier. Between June and August, the DIC was enriched in 13C by 0.7 ‰ on average, while δ18Owater was isotopically heavier by 0.23 ‰ on average (Table 2; Fig. 2). These results agree with studies performed in 2008, when the δ13C in the DIC and the δ18O in Lake Lednica water sampled in August were heavier by 1.2 and 0.35 ‰, respectively, compared with the water sampled in June (Pełechaty et al. unpublished data).

Table 2 Mean arithmetic values of δ13CDIC and δ18Owater in the surface water samples derived along each of the transect established in the littoral zone of Lake Lednica
Full size table

The stable isotope compositions in the Lake Lednica waters differed substantially between 2008 and 2011. The mean δ13C values in the DIC measured in the surface waters in June and August were −5.5 and −3.83 ‰ in 2008 and −6.41 and −5.76 ‰ in 2011, respectively. Accordingly, the stable oxygen isotope values in the lake water were −3.85 and −3.49 ‰ in 2008 and −4.43 and −4.2 ‰ in 2011, respectively. The principal reason for the differences observed is the amount of precipitation, which was distinctly higher in 2011; another reason may be (to a lesser degree) the lower mean summer temperature in 2011.

In June 2011, differences in δ13C in the surface and bottom DIC were absent in most of the samples (Fig. 2). At a few sites, an insignificant enrichment of 0.1-0.17 ‰ in the bottom waters was found. In August, the difference between the surface and bottom δ13C DIC increased with depth, from uniform values at 1 m to 3 ‰ at 6 m, with the bottom waters being isotopically lighter.

In June 2011, uniform δ18O values were measured in the surface and bottom waters (Fig. 2). The differences observed are within analytical error. In August, the δ18O values in the surface and bottom waters were uniform up to 4 m of depth. An insignificant depletion of 18O in the bottom waters of 0.11 ‰ in comparison with the surface waters is noted at 5 and 6 m.

Stable isotope compositions of Dreissena polymorpha shells

Dreissena polymorpha occurring on macrophytes was represented by individuals of different sizes and, hence, different ages. The isotope composition of a shell is thought to reflect the isotope composition of ambient water (Wurster and Patterson 2001), and shells of different sizes have different isotope signatures because they precipitated during different time spans with shifts in the stable isotope composition of water. Thus, special attention was paid to uniform size of the shells chosen for isotope analyses. From a particular transect and depth (Table 1; Fig. 1) (hereafter called the sampling site), 10 D. polymorpha shells with the most uniform size possible were chosen for isotope analyses (Fig. 3).

Fig. 3
figure 3

Stable carbon and oxygen isotope compositions in D. polymorpha shells (filled diamond, open diamond) and bottom water samples (filled circle, open circle) presented separately for each sampling site, i.e., transect and depth. Numbers in the down-right corner of each graph inform about length of the shells analyzed. Filled symbols (filled diamond, filled circle)—snail shells and water sampled in June. Empty symbols (open diamond, open circle)—snail shells and water sampled in August. Mean stable isotope compositions in shells sampled in August (crosses)

Full size image

The stable carbon isotope values ranged between −9.7 and −6.2 ‰ and between −8.9 ‰ and −6.1 ‰ in the D. polymorpha shells sampled in June and August, respectively (Fig. 3). The δ18O values measured in the shells sampled in June were between −4.9 and −1.5 ‰ and between −4.9 and −2.6 ‰ for those sampled in August.

Spread of stable isotope values in Dreissena polymorpha shells derived from one sampling site

The spread of δ13C values in the coeval D. polymorpha shells derived from one sampling site is between 0.6 and 3.1 ‰, with most values falling between 0.7 and 2.7 ‰ (Fig. 3). The smallest spread was observed within the smallest shells, at up to 6 mm long, i.e., 0.6–1.3 ‰. The greatest spread, i.e., 2.1–2.7 ‰, characterised the largest shells, which ranged in length from 12 to 19 mm. The substantial spread in δ13C values was observed even in the samples where the shells were of uniform size, e.g., the δ13C spread of 2.38 ‰ within 10 shells that were all 8 mm long or the 1.37 ‰ spread within 10 shells that were all 9 mm long (Fig. 3).

The spread of δ18O values in the coeval D. polymorpha shells derived from one population was between 0.4 and 2.8 ‰, with most values between 0.7 and 2.3 ‰ (Fig. 3). The smallest spread was observed within the smallest shells, up to 6 mm long, i.e., 0.4–1.4 ‰. The greatest spread, i.e., 1.15–2.76 ‰, characterised the largest shells, at 15–19 mm long; this was not a rule, however, and numerous exceptions were present (Fig. 3). As in the δ13C values, the substantial spread in δ18O values was observed even in the samples where the shells were exactly the same size, e.g., 1.79 ‰ within 10 shells all 8 mm long, 2.29 ‰ with a shell size of 8 mm, 2.05 ‰ at 9 mm and 0.73 ‰ at 2 mm (Fig. 3).

The spread in both the δ13C and δ18O values measured in the individual D. polymorpha shells was smaller, though still significant, when shells sampled in August were considered (Fig. 3).

The above data indicate that the spread of δ13C and δ18O values measured in coeval, individual D. polymorpha shells taken from each of the sampling sites is substantial. The spread increases with the size of the shells; however, it is substantial even within the smallest shells (0.6–1.3 ‰ and 0.4–1.4 ‰ for δ13C and δ18O, respectively) or in shells of exactly the same size (1.4–2.4 ‰ and 0.7–2.3 ‰ for δ13C and δ18O, respectively).

Discussion

Stable isotope composition of the Lake Lednica waters

The uniform stable isotope compositions (δ13C and δ18O) of the surface waters within the lake indicates substantial mixing of the epilimnetic waters (Fig. 2). The approximately 0.2–0.4 ‰ spread observed in the stable carbon isotope composition of the DIC within the samples analysed may result from local differences in the intensity of phytoplankton activity or the wave intensity, i.e., the increased exchange of CO2 with the atmosphere where water turbulence is greater.

The enrichment of surface DIC in 13C by 0.7 ‰ on average observed between June and August (Fig. 2) is interpreted as resulting from the photosynthetic activity of autotrophs. Evaporation is considered the main reason for the 0.2–0.25 ‰ increase in δ18O in the surface waters.

The changes observed in δ13CDIC with depth (Fig. 2) result from the photosynthetic enrichment of surface waters in 13C and the depletion of bottom waters in 13C due to the decomposition of organic matter. Due to spring mixing and, possibly, the early stage of phytoplankton development, this phenomenon was not observed in June; however, in the course of the vegetation season it strengthened. As in the δ13C values, the spring mixing excluded the differences in δ18O between the surface and bottom waters. The insignificant enrichment in 18O (~0.11 ‰) observed between the surface waters and the waters at 5 and 6 m of depth may have resulted from the influence of evaporation.

Spread of the stable isotope composition in the Dreissena polymorpha population taken from one sampling site

The spread in both the δ13C and δ18O signatures in the D. polymorpha shells is significant (Fig. 3). Within one sampling site, all of the shells were collected from the macrophyte patch covering the bottom surface of ~0.02 m2; therefore, the possible spatial difference in the stable isotope composition in the waters is excluded. Even if the area of each of the sampling sites where the macrophytes were collected was greater, the stable isotope spatial diversity in the Lake Lednica waters proved both insignificant and smaller than the range of the δ13C and δ18O values in the shells (Figs. 2, 3).

Although observed in all samples, the spread in isotope values measured in the individual shells was smaller within the smallest shells (Fig. 3). This finding has two implications.

First, it proves that the differences in δ13C and δ18O values observed between the coeval shells of bivalves within one population occur from the early ontogenic stages. Hence, it is suggested that the spread in isotope values recorded in the shells is characteristic of D. polymorpha and results from the bivalve physiology, possibly the control over the stable isotope composition of ions included in the shell structure. Similar observations were made for several species of freshwater snails characterised by a substantial spread in δ13C signatures in coeval shells derived from one population (Apolinarska, unpublished data). Thus, the spread of isotope values in coeval shells occurring within one population is most likely a characteristic feature of molluscs.

In a detailed study of carbon isotope compositions in Margaritifera margaritifera L., Geist et al. (2005) tested whether the δ13C records in shells are controlled by environmental variables. Four nearly 100-year-old mussels exposed to the same environmental conditions had distinctly different and contradictory stable carbon isotope records. The authors explained this phenomenon as resulting from strong metabolic influences on shell precipitation resulting in 13C depletion. Among the factors related to the metabolic activity of mussels that may result in varying body composition and shell carbon signatures between individuals exposed to the same environment, Geist et al. (2005) listed the following: individual age, gender, fecundity and filtering activity.

Second, increased differences in isotope values observed within populations composed of larger (i.e., older) individuals may indicate that the factors controlling the spread in δ13C and δ18O values in the shells enhance their impact with shell growth. However, this may also result from ontogenic differences between individuals. During the shift from the larval to juvenile stage, the mussel attaches itself to hard surfaces by byssal threads (Birnbaum 2011), and if it is not disturbed by physical or chemical conditions, it does not change its habitat. It was observed that populations of D. polymorpha attached to plants were dominated by mussels less than one year old. As the mussel colonies grew, they sank the macrophytes to which they were attached (Stańczykowska and Lewandowski 1993). Subsequently, mussels may move and change their habitat within the lake. In the presented study, both juvenile and adult individuals were present on macrophytes. The shells were present mainly on Charophyte stems, macrophytes that may be one year or perennial. Adult D. polymorpha may also detach their shells and search for a more suitable habitat (Birnbaum 2011). However, as stated above, the spatial difference in the stable isotope composition of the Lake Lednica waters is insignificant and does not explain the significant spread in the stable isotope composition of the shells.

Mean δ13C and δ18O values in Dreissena polymorpha shells versus isotope composition of the water.

The spawning period of D. polymorpha is prolonged and may take place over a period of 3–5 months (Zaiko and Olenin 2006), i.e., between spring and late summer/early autumn. Larvae are found within the plankton between May and September (Stańczykowska 1977). Spawning begins at 12–15 °C and is optimal at 18–20 °C, whereas the optimal temperature for larval development is 20–22 °C (Benson and Raikow 2011). Larvae remain planktonic for up to 4 weeks before settling to the bottom to search for a suitable substratum (Benson and Raikow 2011). D. polymorpha is characterised by rapid growth and may mature within the first year of life under optimal conditions (Benson and Raikow 2011). Molluscs with shells as small as 8 mm may reproduce (Zaiko and Olenin 2006). The optimal temperature for adults extends to 20–25 °C (Benson and Raikow 2011).

Due to the prolonged reproductive period of D. polymorpha within the warm season, it was assumed that all shells sampled in August (length between 2.5 and 11 mm) were spawned in the year of collection (2011), whereas other bivalves, especially individuals exceeding 11 mm, were at least 1 year old. Comparisons of the stable isotope composition of the water with that of shells were performed exclusively for the shells of individuals grown in the year of research and sampled in August. These comparisons were made to exclude larger individuals with shells that had precipitated beyond the period studied.

The exact parameters compared were the mean stable isotope composition of the coeval D. polymorpha shells sampled in August and the mean arithmetic value of the stable isotope composition of the water (June and August values, Table 3), all of which were taken from one sampling site (i.e., a particular transect and depth).

Table 3 Mean δ13C DIC values in waters sampled in June and August compared with mean δ13C values in shells sampled in August
Full size table

δ13Cshell versus δ13CDIC

In the samples where mean δ13Cshell and δ13CDIC were compared, a significant depletion of D. polymorpha shells in 13C was observed (Table 3; Fig. 3). The shells were depleted in 13C by 1.5–2.07 ‰ in comparison with δ13CDIC. Within transect L7, the difference in carbon isotope composition between shells and DIC was highly constant, i.e., −1.65 to −1.69 ‰ (at 1, 4 and 5 m of depth) and 1.89 ‰ (at 3 m of depth) (Table 3). The only exception was noted at 6 m of depth in transect L6, with shells enriched by an average of 0.6 ‰ in 13C relative to δ13CDIC.

The aragonite-HCO3 − fractionation factor was determined experimentally as 2.7 ‰ (Romanek et al. 1992), which means that aragonite shells precipitated in equilibrium with HCO3 − were enriched by 2.7 ‰ in 13C. After subtracting this value from the mean δ13C values recorded in the D. polymorpha shells, the calculated HCO3 − was even more depleted in 13C than in the measured δ13CDIC (Table 3; Fig. 3). Thus, the comparison of the stable carbon isotope composition in the shells and the DIC indicates disequilibrium in their precipitation.

The depletion of 13C in mollusc shells has been reported in numerous studies (Tanaka et al. 1986; McConnaughey et al. 1997; Veinott and Cornett 1998; Kaandorp et al. 2003). The stable carbon isotope composition in D. polymorpha shells was previously studied by Wurster and Patterson (2001); however, δ13CDIC was not measured, so no comparisons with δ13Cshells were possible.

Carbon incorporated into the bivalve shell structure is derived from two sources: ambient DIC and metabolic, i.e., respired CO2 (Dettman et al. 1999; McConnaughey and Gillikin 2008). The ratio of the two sources changes during bivalve ontogeny. Juvenile shells incorporate more carbon from DIC, whereas the contribution of metabolic carbon is increased in the shells of adult individuals (Lorrain et al. 2004; Gillikin et al. 2009; Schöll-Barna et al. 2012). The incorporation of metabolic carbon will deplete 13C in shells relative to δ13CDIC in accordance with the observations in the present study. In a detailed study of carbon isotope compositions in M. margaritifera L., Geist et al. (2005) explained 13C depletion in shells as resulting from strong metabolic influences during shell precipitation. Among the factors related to the metabolic activity of mussels, Geist et al. (2005) listed individual age, gender, fecundity and filtering activity. Geist et al. (2005) also estimated that for shell formation, less than 10 % of the respired CO2 is needed, the amount that allowed the 13C depletion of the shells observed by the authors, assuming a food signature of −27 ‰.

The kinetic isotope effects that occur during rapid calcification are another explanation for the non-equilibrium fractionation and depletion of heavier isotopes in shells (McConnaughey et al. 1997). D. polymorpha is a fast-growing mussel (Stańczykowska 1977), and rapid skeletogenesis can influence the stable isotope composition of its shells. However, the influence of kinetic effects on the stable isotope composition of the shells would result in both 13C and 18O depletion, whereas the shells of the D. polymorpha analysed are usually close to oxygen isotope equilibrium with ambient water or are enriched in 18O relative to water.

δ18Oshell versus δ18Owater

The mean stable oxygen isotope composition of shells is close to the δ18Owater values. The δ18Oshell is between 0.43 ‰ depleted and 0.74 ‰ enriched in 18O relative to δ18Owater (Table 3; Fig. 3). Due to the aragonite composition of D. polymorpha shells (Pathy and Mackie 1993), a 0.6 ‰ enrichment in 18O compared with δ18Owater results directly from the shell mineralogy (Tarutani et al. 1969). After correction for the aragonite composition of the shells, the δ18Oshell–δ18Owater difference ranges between −1.03 to +0.14 ‰.

In general, δ18Oshell is closer to δ18Owater than δ13Cshell is to δ13CDIC. However, to investigate whether D. polymorpha precipitates its aragonitic shell in equilibrium with water, the measured mean δ18Oshell values were compared with those predicted. The equation by Grossman and Ku (1986) for freshwater aragonite was applied:

$$ {\text{T}}\left( {^\circ {\text{C}}} \right) = 20.6 - 4.34\left( {{{\updelta}}^{18} {\text{O}}_{\text{shell}} - {{\updelta}}^{18} {\text{O}}_{\text{H2O}} } \right) $$
(1)

where δ18OH2O is relative to the VSMOW standard subtracted by 0.2 ‰ to relate to the VPDB standard (Table 3).

The calculated δ18Oshell values are close to the mean measured compositions. Both enrichment (by up to 0.27 ‰) and depletion (up to −0.89 ‰) in 18O from the measured values are observed (Fig. 4; Table 3). It is suggested that D. polymorpha precipitates its shells close to δ18O equilibrium with water. However, it must be stressed that the δ18Oshell values under discussion are mean values derived from 10 shells per sampling site. Shells derived from different sampling sites are not uniform in size (Fig. 3), so they represent slightly different δ18Owater values. Moreover, the δ18Owater and water temperature used in the calculations are mean values of only two measurements, made in July and August, for each sampling site.

Fig. 4
figure 4

Mean δ18O shell values in D. polymorpha measured in shells sampled in August and calculated shell oxygen isotope compositions using equation of Grossman and Ku (1986) with mean temperature measured in June and August and mean δ18Owater values sampled in June and August

Full size image

The above isotope comparisons provide a general water-shell relation. Direct, detailed comparisons would require a more detailed study of both the shell (and, possibly, shell layers) and the water sampled in closer intervals during the season of shell growth.

Conclusions

The findings of the present paper may contribute to the investigation of stable isotope compositions of both recent and ancient environments dominated by D. polymorpha. It is essential to note that, according to the results obtained in the present study, the stable isotope composition of a single shell does not reflect the carbon and oxygen isotope compositions of the ambient water. Shells of coeval bivalves from one population sampled at one restricted site may differ by as much as 3.1 and 2.8 ‰ when the δ13C and δ18O values of the shells are considered, respectively. Thus, isotope investigations of D. polymorpha shells should not be performed on single shells, as the isotope values will not be representative of the population.

The depletion of shells in 13C relative to DIC is consistent with other investigations of freshwater molluscs. It is suggested that this depletion is of metabolic origin. Because the disequilibrium between δ13Cshell in D. polymorpha and δ13CDIC observed in the recent populations studied is constant, i.e., between 1.5 and 2 ‰, this proxy may be successfully used in palaeoenvironmental studies of stratigraphic sequences.

The mean δ18O values derived from ten coeval shells of D. polymorpha at particular sampling sites are close to the calculated stable oxygen isotope compositions of the shells. Thus, it is suggested that the δ18O in D. polymorpha shells may be used as palaeoclimatic and palaeoenvironmental proxy. However, due to the spread in δ18O values observed, samples for isotope analyses should be composed of several shells, e.g., ten, as in the present study.

The major conclusion of the study, that isotope analyses of single shells are not useful in paleolimnological investigations, agrees with previous findings by Jones et al. (2002) who studied stable carbon and oxygen isotope compositions of two gastropod shells Gyraulus piscinarium and Valvata cristata. Similar conclusions were made also by Escobar et al. (2010) based on ostracod shells. However, the two studies dealt with subfossil material derived from 1-cm stratigraphic sediment sections, whereas in the present paper it was shown that the stable isotope composition of shells changes not only in time but also among coeval molluscs of one species inhabiting exactly the same environment at the same time.