5. DISCUSSION
5.1 Factors affecting δ13CDIC in the Genggahai Basin waterbodies
Several major processes affect the stable carbon isotope compositions of the lake water DIC, such as in-lake processes (including lake metabolism, organic matter decomposition, calcite precipitation, and exchange with atmospheric CO2) and the climatic and geographical environment of the catchment (including carbonate rock weathering and dissolution and soil respiration) (Bade et al., 2004). The climatic and geographical environment in the catchment can alter the carbon isotope composition of the lake water DIC by influencing the aqueous CO2 and alkalinity of inflowing water. Generally, the five following factors affect the isotope composition of the DIC.
1) Lake inflow carbon isotope composition. The isotopic composition of the lake inflow directly affects the isotopic composition of the lake. This is especially significant in exorheic lakes or lakes with a short water retention time. In the Chara spp. community, the δ13CDIC-L and δ13CDIC-I values were positively correlated (r 2 = 0.68, p < 0.01), whereas in the M. spicatum community, these values were uncorrelated (Fig. 5). This is because the M. spicatum community was located far from the lake inflow in the south-eastern region, where exchange with spring water was weak (Fig. 1b). This reveals that the δ13CDIC-I values were key factors influencing the δ13CDIC-L values.
{Figure 5}
2) Exchange with atmospheric CO2. The DIC pool in the lake tended to be in isotopic equilibrium with the atmosphere via CO2 exchange. During the CO2 exchange process, 12C-rich CO2 is preferentially released from the lake surface to the atmosphere, yielding a DIC pool enriched in 13C. This exchange process is slow; therefore, its effect on the δ13CDIC-L values is more notable in endorheic lakes characterised by long retention time, whereas it is not as observable in lakes with short retention times or rapid circulation (Shen et al., 2010). When lake water to atmospheric CO2exchange reaches an equilibrium, the δ13CDIC-L values range from 1–3 ‰ (Deuser and Degens, 1967; Leng & Marshall, 2004).
The exchange between the DIC of lake water and atmospheric CO2 is continuous in exorheic lakes. At equilibrium, isotope fractionation occurs between atmospheric CO2 and dissolved carbonate species, i.e., CO2 (aq), HCO3, and CO32– (Zhang et al., 1995), as follows:
ε aq-g = –(0.0049 ± 0.003) × T (℃) – (1.31 ± 0.06), (1)
ε HCO3-g = –(0.141 ± 0.003) × T (℃) + (10.78 ± 0.05), and (2)
ε CO3-g = –(0.052 ± 0.03) × T (℃) + (7.22 ± 0.46) (3)
Temperatures recorded using a water level data logger showed that from May to September of 2013–2015, the mean water surface temperature of the Genggahai Lake was 17.0 ℃. Based on Eqs. (1)–(3), the carbon isotope fractionation factors between H2CO3, HCO3, and CO32– and atmospheric CO2 were –1.5, 8.38, and 5.89 ‰, respectively. The isotopic composition of global atmospheric CO2 is approximately –8.1 ‰ (Das et al., 2005). Therefore, at equilibrium, the isotopic values of H2CO3, HCO3, and CO32– in lake water are –9.6, 0.28, and –2.21 ‰, respectively. Different forms of DIC have varying δ13C values at isotopic equilibrium; the magnitude of the δ13CDIC value depends on the proportions of the different DIC forms in lake water, which is related to its pH value (Stumm & Morgan, 1970). When the pH value was 5.5, 80 % of the DIC in a water body has an aqueous CO2 form (aq). When the pH is 8.5, CO2 (aq) accounts for < 1 % of the DIC, which predominantly takes HCO3 and CO32– forms. When the pH reaches 10, HCO3 accounts for < 50 % of the DIC, whereas CO32– dominates the DIC (Stumm & Morgan, 2012). The pH value of the Genggahai Lake varied from 8.1 to 10.6, which indicates that HCO3 and CO32– were its dominant forms of DIC. Notably, the actual δ13CDIC values of the Genggahai Lake were generally more negative than the atmosphere-equilibrated δ13CDIC.
3) Organic matter decomposition in lake sediments. Sedimentary organic matter in lakes includes native aquatic plants and terrestrial organic debris transported into the lake from the surrounding watershed. Once degraded, this organic matter increases the12C-enriched DIC composition of lake water (Myrbo & Shapely, 2006). Organic matter decomposition in Qingmuke lake (i.e., a freshwater lake located on the Qiangtang Plateau) resulted in a DIC isotope value equal to or even lower than that of river water (Lei et al., 2012). In contrast, the isotopic composition of the DIC in the Genggahai Lake was significantly more positive than that of the Shazhuyu River, indicating that organic matter decomposition may have had a relatively small effect on the DIC composition of the lake. Additionally, methane produced by organic matter decomposition resulted in a more negative δ13CDIC value. Organic matter decomposition can cause a decrease in the δ13CDIC values to –50 ‰ (Sun et al., 2013). This value is significantly lower than the mean δ13CDIC-L value of the Genggahai Lake, which indicates that the CO2 or methane produced via decomposition did not have a significant effect on seasonal or interannual changes in the δ13CDIC-Lvalues.
During the observation period, the mean carbon isotopic composition of organic matter (δ13Corg) in theChara spp. community was –16.0 ‰, whereas the mean values of δ13Corg in the P. pectinatusand M. spicatum communities were –12.7 and –11.4 ‰, respectively. If we neglect the effect of carbon isotope fractionation owing to organic matter decomposition, the δ13C of the CO2 released via organic matter decomposition is then equal to δ13Corg. According to Eq. (2), if HCO3 is the dominant form of DIC in the lake, then the equilibrium isotopic value of HCO3− in the lake water is 0.28 ‰. In this case, the mean δ13CDIC-L value of theChara spp., P. pectinatus , and M. spicatumcommunities would be –15.72, –12.42, and –11.12 ‰, respectively. However, the observed mean δ13CDIC-Lvalues for these three communities were –5.4, –7.4, and –7.9 ‰, confirming that organic matter decomposition has a limited effect on the δ13CDIC-L of the Genggahai Lake.
4) Lake photosynthetic activity. In highly productive lakes, photosynthesis is a key factor that affects the δ13CDIC values of the lake water (McKenzie, 1982, 1985). During photosynthesis, plants preferentially uptake 12C, which yields more negative δ13C values for plants and the δ13CDIC of the water body becomes more positive (Andrews et al., 2004). Charaphytes are an important submerged aquatic macrophyte. Compared with vascular plants, charaphytes have a higher photosynthetic rate and lower respiration rate. The preferential uptake of 12CO2 for photosynthetic purposes could have led to the 13C-enrichment of the DIC in the lake water (Pełechaty et al., 2010). During intense photosynthesis, dissolved CO2 in lake water is limited (Herczeg & Fairbanks 1987). When this occurs, charaphytes use HCO3 for photosynthetic activity. Compared with vascular plants, charaphytes can use HCO3 for photosynthetic activity more effectively (van den Berg et al., 1999). According to Eqs. (1)–(3), the δ13C values of HCO3 were more positive than those of H2CO3 and CO32– in the lake water. In contrast, the photosynthetic activity of charaphytes results in carbonate precipitation in the surrounding waters, forming thick CaCO3 encrustations (Pełechaty et al., 2010). This also led to 13C-enriched water in the charaphyte growth area.
We found that the seasonal bias in the δ13CDIC-L values of the Charaspp. community were more positive in July (2012–2014) (Fig. 4a). Additionally, at the beginning of Chara spp. growth (in May), the δ13CDIC-L value of the Charaspp. community was equal to the value at the end of Chara spp. growth (in September), especially in 2012. In contrast, the δ13CDIC-L values in the P. pectinatus and M. spicatum communities showed no seasonality (Fig. 4a). This phenomenon may have occurred because Chara spp. has a higher photosynthetic rate and a lower respiration rate than that of the other communities (Van den Berg et al., 2002). Pentecost et al. (2006) also observed seasonal variations in the δ13CDIC-L values of the Charaspp. community in the UK. Pełechaty et al. (2010) suggested that an increase in δ13CDIC results from the intense photosynthetic activity of Chara rudis during the early summer. Moreover, we found that the differences in the δ13CDIC-L values between Charaspp. and vascular plants were smaller at the beginning and end of the growing season in the Genggahai Lake, but larger during the mid-growth season (July) (Fig. 4a). There may have been a limited impact from photosynthesis on the δ13CDIC-L values in areas with submerged vascular plants. This trend was not evident during certain months, e.g., in July 2015. Due to data limitations, we cannot provide an explanation for this phenomenon. Nevertheless, we can reasonably conclude that the variations in the δ13CDIC-L values of the lake water were related to the intensity of photosynthetic activity in different aquatic plants.
5) Water retention time. In arid regions, with extended lake water residence times, strong evaporation leads to the preferential loss of the lighter 12CO2 and16O2 isotopes, yielding more positive δ13CDIC-L and oxygen (δ18O) isotopic lake water compositions; furthermore, there was a significant positive correlation between δ13CDIC-L and δ18OL (Li & Ku, 1997). Monitoring results revealed that the δ18OL values of the Genggahai Lake deviated significantly from the global meteoric water line, but were consistent with the local evaporation line (LEL), indicating that evaporation affected the δ18OL composition of the lake water (Jin et al., 2015; Qiang et al., 2016). However, this study found that the δ13CDIC-L and δ18OL values of the Genggahai Lake were not correlated (Fig. 6), which indicates that evaporation may have had only a minimal effect on the δ13CDIC-L value of the lake.
{Figure 6}
The preceding analysis demonstrated that the lake surface to atmospheric exchange of CO2 and evaporation had a relatively minimal effect on δ13CDIC-L; δ13CDIC-I primarily influenced the changes in δ13CDIC-L. The high photosynthetic efficiency of Chara spp. indicates that the its corresponding δ13CDIC-L values showed a seasonal trend of more positive values, which were more positive than the δ13CDIC-L values of areas with vascular plants. This shows that the photosynthetic activity of vascular plants has a negligible effect on the δ13CDIC-L of lake water.
5.2 Isotopic composition of DIC in the Genggahai Basin groundwater and Shazhuyu River
The δ13CDIC-I and δ13CDIC-R values were significantly more negative than δ13CDIC-L during the same period, which suggests that isotopic fractionation owing to atmospheric exchange or the photosynthesis of aquatic plants occurred after groundwater inflow into the lake (Leng & Marshall, 2004). Compared with the δ13CDIC-I value, δ13CDIC-R was significantly more positive, possibly because the Shazhuyu River essentially forms via surface runoff, as well as the more frequent exchange of atmospheric CO2 with surface water than groundwater.
We found a significant correlation between δ13CDIC-I and δ13CDIC-R (n = 17,r2 = 0.46, and p < 0.01). Additionally, given that groundwater is the main water source of the Genggahai Lake, δ13CDIC-I was the main factor affecting the δ13CDIC-L values of the lake water (see section 5.1). Therefore, we further analysed the influencing factors of δ13CDIC-I.
Three main species compose the DIC in water bodies: CO2, CO32–, and HCO3. In this study, the titration method was used to determine the DIC composition of groundwater in the Genggahai Basin. We found that HCO3was the dominant form of DIC. Additional research has shown that the primary source of HCO3 in the groundwater of the Genggahai Basin and the Shazhuyu River is the chemical weathering of rocks, especially carbonate rocks (Jin et al., 2019). Based on the reaction equation for chemical weathering, CO2 is an essential component of this process. Dissolved CO2 in groundwater originates from the atmospheric flux, watershed soil respiration, and organic matter decomposition (Gu, 2011). Atmospheric CO2 normally has δ13C values of approximately –8 ‰ (Leng & Marshall, 2004), which is relatively stable. The δ13C values of soil CO2 from areas with C3 and C4 plants range from –32 to –20 ‰ (Brunet et al., 2005) and –17 to –19 ‰, respectively. The δ13C value of HCO3 from carbonate dissolution during subsurface weathering is approximately 0 ‰ (Weynell et al., 2016). The isotopic equilibrium between CO2 and HCO3 from different sources can be attained through the following exchange reaction:
13CO2(g) + H212CO3(aq) =12CO2(g) + H213CO3(aq). (4)
The fractionation factor of CO2 and HCO3 in the CO2–HCO3 system in soil is approximately 10 ‰ (Leng & Marshall, 2004). After soil CO2 is dissolved in water, its δ13C value becomes more negative than that of HCO3. However, carbonate rock dissolution increases the δ13C value of HCO3. Evidently, the carbon source that affects the isotopic composition of DIC in groundwater, δ13CDIC-I, consists of two components: (1) 13C from the weathering and dissolution of carbonate rocks, which possesses a more positive isotope ratio and (2)12C from CO2 generated through soil respiration, which has a more negative isotope ratio. Therefore, the relative contribution of these two carbon sources to the groundwater DIC determines the composition of δ13CDIC-I in the Genggahai Basin.