Carbon isotopic records of middle Holocene corals and environmental causes in northern South China Sea

Quaternary International 349 (2014) 257e269

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

Carbon isotopic records of middle Holocene corals and environmental causes in northern South China Sea Ruixia Su a, *, Yechun Zhuang a, b, Dandan Sui c, Yufen Xu b, Lizhuo Sun d, Hai Chen e, Donghuai Sun f, Baofeng Li f, Fei Wang f a

Key Laboratory of Marginal Sea Geology, South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 West Xinggang Road, Guangzhou, 510301, China Center of Environment Engineering, South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 West Xinggang Road, Guangzhou, 510301, China c Key Laboratory of Tropical Marine Environmental Dynamics, South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 West Xinggang Road, Guangzhou, 510301, China d Department of Biotechnology, College of Life Science of Yangtze University, 88 Jingmi Road, Jingzhou, 434023, Hubei Province, China e Minnesota Isotope Laboratory, Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN, 55455, USA f Key Laboratory of West China's Environmental System (Ministry of Education), Lanzhou University, 222 South Tianshui Road, Lanzhou, 730000, China b

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 7 September 2014

Two high-resolution d13C records of ~5.4 and ~4.4 ka coral colonies Porites lutea from east coast of Hainan Island in the northern South China Sea (SCS) provide insights into the complex links between coral carbon isotopes and environment changes. The d13C of the ~5.4 ka coral offered a 40-year growth history with an average of 2.62‰ and amplitude of 2.96‰, the ~4.4 ka coral's revealed a 55-year growth history with an average of 3.12‰ and amplitude of 4.21‰, which revealed higher coral d13C and increased seasonality in the middle Holocene. Our analysis suggests that the higher coral d13C was the combined result of higher radiation, salinity and nutrient level and lower atmosphere CO2 concentration, which might increase coral d13C through strengthening photosynthesis of symbiont zooxanthella in the context of kinetic and metabolic fractionations. The increased seasonality was mostly produced by strengthened radiation. The decreasing standard deviation of the ~5.4 ka coral annual d13C revealed weakened interannual changes of atmospheric CO2, salinity, and nutrient. By contrast, the increasing deviation of the ~4.4 ka coral d13C suggests strengthened interannual changes of these variables. The ~5.4 ka coral d13C showed a long-term increasing trend at a rate of 0.33% increase y 1, which was attributed to natural decrease in atmospheric CO2, and increasing salinity and nutrient level. However, the ~4.4 ka coral d13C showed a long term decreasing trend at a rate of 0.25% decrease y 1, which was ascribed to unusual increase in atmospheric CO2, and decreasing salinity and nutrient level. © 2014 Elsevier Ltd and INQUA. All rights reserved.

Keywords: Coral d13C Solar radiation Atmospheric CO2 concentration Salinity Holocene South China Sea

1. Introduction Keith and Weber (1965) first found that unlike other organisms, corals frequently form their skeletons out of “isotopic equilibrium” with their environment. The isotopic disequilibrium is attributed to the combined effect of kinetic and metabolic fractionation (McConnaughey, 1989a, b). Kinetic fractionation is associated with CO2 hydration and hydroxylation, and it results in depletion of 13C in carbonate. In this model, the coral pumps protons from the

* Corresponding author. E-mail address: [email protected] (R. Su). http://dx.doi.org/10.1016/j.quaint.2014.08.030 1040-6182/© 2014 Elsevier Ltd and INQUA. All rights reserved.

region in the skeleton forming raising the pH and lowering pCO2. Molecular CO2 then diffuses across the coral basal epithelium and reacts with H2O and OH to produce HCO3 and CO23 . This raises the CaCO3 saturation state, and allows for rapid calcification. The starting CO2 has relatively low 13C content. Kinetic discrimination against the heavy isotopes during CO2 reactions further lowers the 13 C content of product HCO3 . During fast calcification, this isotopically depleted HCO3 precipitates before it can isotopically equilibrate with seawater dissolved inorganic carbon (DIC). In contrast, the effect of the metabolic fractionation on coral d13C is more complicated (Swart, 1983; McConnaughey, 1989a, b). According to the common carbon pool assumption, photosynthesis preferentially uses 12C, leaves the DIC pool enriched in 13C (Goreau, 1977;

258

R. Su et al. / Quaternary International 349 (2014) 257e269

Swart, 1983). As the metabolic demand for carbon increases with coral growth, the pool of available 12CO2 in the coral becomes relatively depleted causing increase to the d13C signature of coral (Erez, 1978; Swart, 1983; McConnaughey, 1989b). As photosynthesis is the faster reaction in light, the reservoir enriched in 13C caused a d13C increase with light (Goreau, 1977). On the contrary, during respiration of organic matter, addition of 12C-enriched CO2 decreases the d13C of the skeletal carbonate (Smith and Kroopnick, 1981) due to reduction in photosynthesis and decrease in metabolic processes (Grottoli et al., 2004). Factors such as cloud cover can play a key role in carbon assimilation of symbiotic corals (Fairbanks and Dodge, 1979; Quinn et al., 1993; Sun et al., 2008) where d13C signature has been shown to be a recorder of light level or solar radiation due to radiation induced changes in metabolic rate (Coles and Jokiel, 1978; Swart, 1983; McConnaughey, 1989a; Grottoli and Wellington, 1999), such as, moderate increase in the rate of photosynthesis, related to increase in light intensity, appears to increase the coral skeleton d13C, while decrease in light results in low d13C in the skeleton (Swart et al., 2005). Obviously, metabolic effects are responsible for large shifts of skeletal d13C, which results primarily from the seasonally variant ambient light incident on the coral surface (Coles and Jokiel, 1978; Swart et al., 2005, 2010). However, characteristics of interannual variations in coral d13C are difficultly explained by solar radiation, but easily interpreted by other factors, such as, sea surface salinity (SSS) indicating seawater components related to terrestrial runoff and rainfall (Swart et al., 1996; Su et al., 2007; Craig et al., 2010), nutrient (Grottoli and ~ o-SouthWellington, 1999), spawning (Gagan et al., 1996), El Nin ern Oscillation (ENSO) event and blenching (Grottoli et al., 2004). Coral d13C shows a significant positive response to salinity (Pretet et al., 2014), which is related to precipitation and river discharge related DIC in the coastal ocean (Grove et al., 2012). Moreover, seawater zooplankton, composition, and DIC are used by coral as food and calcification (Gupta et al., 2006), because on the one hand, coral is autotrophic and depend on its symbiont for nutrition, on the other it is heterotrophic and depend on plankton predation (Levy et al., 2006). Coral d13C increases as abundant nutrient and rich food can strengthen the both algal photosynthesis and coral calcification (Grottoli and Wellington, 1999). Thus, d13C reflects the different DIC proportions of calcified CO2 sources from seawater and coral respiration (Rollion-Bard et al., 2003), which may be independent of illuminations (Furla et al., 2000). In nonzooxanthellate coral, the fractionation of the stable isotopes is controlled by the pH of the calcification fluid, as which decreases the skeletal d13C decreases (Adkins et al., 2003). Significantly more negative d13C explains that coral may respire a significant proportion of isotopically negative substances, such as lipids (Swart et al., 2005). Obviously, d13C apparent disequilibrium in coral is the result of a mixing among metabolic CO2 related with solar radiation, DIC, salinity, nutrient and more other factors coming directly from seawater. In addition to the seasonal and interannual variation in coral d13C, several workers have remarked upon long term trend towards lower d13C value within modern coral skeleton, which is attributed to the 13C Suess effect due to the addition of anthropogenically derived CO2 to the atmosphere (Nozaki et al., 1978; Swart et al., 2010). Coral d13C has a positive correlation with pH as CO2 increases in atmosphere and seawater (Sabine et al., 2004), which is closely linked to the anthropogenic CO2 emissions from burning of fossil fuels (Hemming et al., 1998). Coral d13C may also identify long term processes of coral growth under high nutrient loads and potential disturbances, suggesting that mariculture and other human related stressors influence on the carbon fractionation of coral skeleton (Levy et al., 2010). Therefore, distinguishing and

evaluating various processes and environmental variables are necessary if coral d13C values are used to reconstruct marine environments. In this work, the two fossil coral colonies Porites lutea OC1 and OC2, and modern coral P. lutea QG5 were collected at east coast of Hainan Island in the northern South China Sea (SCS), a region well positioned to monitor the Asian monsoon. Previous works have discussed the evolution of Asian monsoon variability in the middle Holocene revealed by the d18O records of the two fossil corals (Sun et al., 2005; Su et al., 2010), based on the researches of relationships between the modern coral skeletal d18O and monsoonal sea surface temperature (SST), SSS and fresh-water input (Su et al., 2006). It has been studied for associations between the modern coral calcification and d13C and solar radiation and cloud cover, climate warming due to increasing anthropogenic CO2 emissions and human activities (Sun et al., 2008), and carbon isotope compositions of SSS indicating seawater inorganic and organic carbon components related to rainfall and freshwater inputs, and increasing atmospheric CO2 concentration (Su et al., 2007). In contrast, the associations between the Holocene coral d13C and environment changes are poorly understood. In this study, the paper offers 40 and 55 year high resolution records of the two fossil coral d13C, which are compared with a 17 year record of the modern coral QG5 d13C. Consequently, Holocene environment and climate causes for coral d13C variability are examined and evaluated at seasonal, interannual and interdecadal scales. 2. Methods 2.1. Study site The corals are located on eastern Hainan Island of the northern SCS (Fig. 1A), strongly affected by the Asian monsoon and northwestern Pacific Ocean. During the summer monsoonal seasons, wetter climate prevails in the area, with higher temperature (25e29  C), lower salinity (4.1e33.7‰), and more plentiful rainfall (2000e2400 mm). During winter seasons, drier climate prevails with lower temperature (16e24  C), higher salinity (20.5e35.5‰), and lower rainfall (1000e1200 mm). Sunshine is from 1750 to 2650 h annually, and the total annual radiation is about 4500e5900 MJ m 2. Detailed millennial climate changes in geomorphology of the Holocene are appropriate for investigation, because periodic uplifts elevated a succession of fossil reefs above present sea level at Hainan Island. Abundant Holocene coral reefs have developed along the coastal shallow sea area around the northern SCS (Sun et al., 2005; Su et al., 2010). 2.2. Coral sampling In February 2002, the two well-preserved fossil corals P. lutea OC1 (Su et al., 2010) and OC2 (Sun et al., 2005) were collected from an emergent Holocene coastal terrace at Oucun, eastern Hainan Island, ~10 km northeastern of Qionghai (Fig. 1A and B). A modern coral P. lutea QG5, growing in a similar reef environmental setting, was collected from the same platform on 5 May, 2002, ~5 km southwest of the paleo-reefs (Fig. 1B; Su et al., 2006). Tetragonal prism samples of the two fossil corals and one modern coral with cross-sections as squares with sides 30 cm were cut in parallel to their main growth axes. 2.3. Chronology and isotopic analysis In the laboratory, the corals were cut into 5-mm-thick slabs in parallel with their main growth axes using an automatic stonecutter, as with the modern coral (Su et al., 2006). After air-drying,

R. Su et al. / Quaternary International 349 (2014) 257e269

Fig. 1. Map of sample site, Hainan Island and SCS. A. Map of study area. B. Regional map of sample sites.

259

260

R. Su et al. / Quaternary International 349 (2014) 257e269

Fig. 2. High-resolution d13C and d18O records of middle Holocene corals. A. d13C and d18O of ~5.4 ka coral OC1; B. d13C and d18O of ~4.4 ka coral OC2.

the coral slabs were photographed using a medical HITACHI X-ray machine to obtain negative X-ray images, which were converted to positive digital images using a UMAX scanner. Then, the annual density banding couplets, composed of high and low-density banding, can be clearly observed. The dates of the two fossil corals were ~5.4 and ~4.4 ka BP, respectively, from the 230Th dating analysis results (Sun et al., 2005; Su et al., 2010). The carbon and oxygen isotopic compositions of the two fossil corals were analyzed on a MAT252 mass spectrometer coupled with an auto-carbonate system (Sun et al., 2005; Su et al., 2010). In total, 458 and 799 micro-subsamples were analyzed for the stable carbon isotopes of the fossil coral OC1 (Fig. 2A) and OC2 (Fig. 2B), respectively, and compared with the modern coral's (Fig. 3A). 2.4. Oceanographic and climatologic data Oceanographic and climatologic data are shown in Fig. 3BeD. Solar radiation is the downward solar radiation flux at sea surface (20 N, 110.625 E), derived from the satellite data from http://www. cdc.noaa.gov/cdc/reanalysis/with a 2.5 resolution, and cloud cover was determined at Qionghai meteorological station (Fig. 3B). Atmospheric CO2 concentration with d13C at Mauna Loa (ML) Observatory of Hawaii was obtained from http://cdiac.ornl.gov/pub/ ndp001/maunaloa.co2 (Fig. 3C; Keeling et al., 2009). SSS is measured daily at 1 m water depth at Qinglan oceanic station, ~40 km northeast of the sampling site (Fig. 3D). The relationship between SSS and local rainfall was determined using daily rainfall recorded at Qionghai meteorological station (Fig. 3D), ~10 km southwest of the coral sampling site. 3. Results 3.1. ~5.4 ka coral d13C The d13C time series of the ~5.4 ka coral displays regular seasonal fluctuations in the form of a clear annual cycle with a superimposed

distinct trend. Sharp winter concave valleys and narrow summer convex apices with bi-peaks follow annual cycles (Fig. 2A). The d13C minimum occurs in SeptembereNovember, ~2e3 months earlier than the d18O maximum in DecembereJanuary (Su et al., 2010). In contrast, the d13C maximum occurs in MarcheMay, being ~2e3 months earlier than the d18O minimum in JulyeAugust. These extreme distributing characteristics indicate that seasonal and interannual change of the fossil coral d13C is as same as the modern coral's, but different from their d18O (Figs. 2 and 3A). The d13C of the ~5.4 ka coral averaged 2.62‰ with a range of 0.97‰, which increased 0.56‰ (18%) higher with a range decrease of 0.04‰ (4%) than the modern coral's; the maximal d13C was 1.18‰ with a range of 1.41‰, which increased 0.87‰ (42%) higher with a range increase of 0.20‰ (17%). The minimal d13C was 4.14‰ with a range of 1.61‰, which increased 0.42‰ (9%) higher with a range increase of 0.36‰ (29%) (Figs. 2e3A; Table 1). These data show that all the average, maximum and minimum of the ~5.4 ka coral d13C were higher, while the ranges of the maximum and minimum were higher, but that of the average was slightly lower. The amplitude between the d13C maximum and minimum was 2.96‰, being 0.42‰ (18%) higher than the modern coral's, but 35% lower than the 4.6‰ of the ca. 4e7 ka coral from Belize, Central America (Gischler and Storz, 2009). Table 1 Averages, maxima, minima and standard deviations (ss) of two middle Holocene coral d13C and their variations relative to modern coral. Coral d13C (‰)

5.4 ka

4.4 ka

Modern

Average

Average Variation Variation (%) Beginning End Variation Variation rate (% yr 1)

2.62 0.56 18 2.79 2.44 0.35 0.33

3.12 0.06 2 2.92 3.32 0.40 0.25

3.18

s

0.21

0.38

2.81 3.53 0.72 1.71 0.30

R. Su et al. / Quaternary International 349 (2014) 257e269 Table 1 (continued ) Coral d13C (‰) Variation Variation (%) Range Variation Variation (%) Maximum

Maximum Variation Variation (%)

s Variation Variation (%) Range Variation Variation (%) Minimum

4.4 ka

Modern

0.09 30 0.97 ( 3.11~ 2.14) 0.04 4

0.08 27 1.61 ( 3.94~ 2.34) 0.60 59

1.01 ( 3.66~ 2.65)

1.18 0.87 42 0.33 0.01 3 1.41 ( 1.83~ 0.42) 0.20 17

0.98 1.07 52 0.65 0.33 103 2.93 ( 2.70~0.23) 1.72 142 5.19 0.63 14

2.05

0.32

1.21 ( 2.95~ 1.74)

4.56

range of 2.93‰, 1.07‰ (52%) higher with a range increase of 1.72‰ (142%). The minimal d13C was 5.19‰ with a range of 1.91‰, 0.62‰ (14%) lower with a range increase of 0.66‰ (53%) (Figs. 2B and 3A; Table 1). These data show that both the average and maximum of the fossil coral d13C were higher, but its minimum was lower with all higher ranges. The amplitude between the maximum and minimum was 4.21‰, being 1.70‰ (68%) higher than the modern coral's, but 0.39‰ (9%) lower than the ca. 4e7 ka coral's from Belize (Gischler and Storz, 2009). The annual average, maximal, and minimal d13C ss of the ~4.4 ka coral are 0.38‰, 0.65‰ and 0.46‰, respectively, being 27%, 103% and 10% higher than the modern coral's (Table 1). These data indicate that interannual variability in coral d13C intensified throughout the year, especially in summer at ~4.4 ka BP. In addition, the annual d13C of the ~4.4 ka coral shows a longterm decreasing trend with a decrease of 0.40‰, from 2.92‰ at the beginning to 3.32‰ at the end, at a rate of 0.26% decrease y 1 over a period of 55 years. These data show that d13C trend variability of the ~4.4 ka coral is consistent with the modern coral's, but its decrease rate was much smaller (Figs. 2B and 3A; Table 1).

Minimum Variation Variation (%)

4.14 0.42 9

s

0.46 0.04 10 1.91 ( 5.94~ 4.03) 0.66 53

0.42

4. Discussion

1.25 ( 5.15~ 3.89)

4.1. Seasonal variability in coral d13C and reconstruction of solar radiation in middle Holocene

Variation Variation (%)

0.41 0.01 2 1.61 ( 5.01~ 3.40) 0.36 29

Amplitude Variation Variation (%)

2.96 0.45 18

4.21 1.70 68

2.51

Variation Variation (%) Range

Amplitude

5.4 ka

As an indicator of interannual variability, the standard deviation (s) of the ~5.4 ka coral annual d13C is 0.21‰, 30% lower than the modern coral's (Table 1), implying that interannual variation in coral d13C probably weakened year-round at ~5.4 ka BP. The s of the annual minimal d13C is 0.41‰, slightly decreased (2% lower), while the annual maximal d13C's is 0.33‰, slightly increased (3% higher). These data show that interannual changes of coral d13C in winter and summer at ~5.4 ka BP were slightly lower and higher, respectively. In addition, the annual d13C of the ~5.4 ka coral showed a longterm increasing trend with an increase of 0.35‰, from 2.79‰ at the beginning to 2.44‰ at the end, at a rate of 0.33% increase y 1 over a period of 40 years (Fig. 2A; Table 1). This indicates that the d13C trend variability of the fossil coral was opposite to the modern's. Moreover, its increase rate was much smaller than the decrease rate of the modern's (Fig. 3A). 3.2. ~4.4 ka coral d13C Similar to the ~5.4 ka (Fig. 2A) and modern coral (Fig. 3A), the

d13C time series of the ~4.4 ka coral displays regular seasonal

fluctuations in the form of clear annual cycle superimposed distinct trend, and sharp winter concave valleys and narrow summer convex apices with bi-peaks follow annual cycles (Fig. 2B). The d13C minimum distributes in SeptembereNovember, being ~2e3 months earlier than the d18O maximum in DecembereJanuary. By contrast, the d13C maximum distributes in AprileJune, being ~2e3 months earlier than the d18O minimum in JulyeAugust. These extreme distributing characteristics show that seasonal and interannual change of the ~4.4 ka coral d13C is as same as the modern and ~5.4 ka coral's but different from their d18O's (Figs. 2 and 3A). The average d13C of the ~4.4 ka coral was 3.12‰ with a range of 1.61‰, 0.06‰ (2%) higher with a range decrease of 0.60‰ (59%) than the modern coral's. The maximal d13C was 0.98‰ with a

261

4.1.1. Seasonal variability in modern coral d13C and solar radiation In order to study seasonal features and variability, the high resolution coral d13C and solar radiation were translated to detrending data (Fig. 4A). Systematic investigations indicated that seasonal variability in modern coral d13C is related to solar radiation (Su et al., 2007; Sun et al., 2008), which has been investigated by analyzing effects of seawater depth (Weber et al., 1976; Grottoli and Wellington, 1999), cloud cover (Quinn et al., 1993), and seasonality (Fairbanks and Dodge, 1979; McConnaughey, 1989a). Comparatively, the modern coral d13C variation is very close to the seasonal pattern of the radiation (Fig. 4A). Further linear regression analysis indicates a good positive correlation with a coefficient of 0.72 (Fig. 4B; Sun et al., 2008). The association between coral d13C and radiation implies that increasing radiation may strengthen symbiont photosynthesis, even though corals grow at rather shallow depths of 3e4 m, which seems to support the model of Goreau (1977). According to the regression equation (Fig. 4B), the reconstructed solar radiation is 178.88 W/m2, close to the 178.89 W/m2 for the solar radiation from satellite data (Fig. 4A), implying that solar radiation is well reconstructed by coral d13C. Moreover, the maximum and minimum of the modern coral d13C consistently occur in AprileJune and NovembereDecember, respectively (Fig. 3A), indicating a strong seasonal offset between d13C and d18O. These suggest that the “kinetic” offset from equilibrium for both 13C and 18O appears to be fairly constant, because the kinetic effect generally results in the simultaneous fractionation of carbon and oxygen isotopes (Sun et al., 2008). The variation in skeleton d13C therefore appears to be driven mainly by ‘metabolic’ effect. Accordingly, seasonal variability in coral skeletal d13C is mainly correlated to seasonal changes in radiation modulation of coral photosynthesis (Quinn et al., 1993). Light intensity, through the activity of the coral's endosymbiotic algae, regulates the depth-dependent and seasonal variation in the skeletal carbon isotopic composition (Fairbanks and Dodge, 1979). Decrease in light or increase in zooplankton or simultaneous decreases in both may result in significant decrease in coral d13C level (Grottoli and Wellington, 1999). Thus, lightinducing photosynthesis has a dominant role on the seasonal change of the modern coral d13C in the northern SCS.

262

R. Su et al. / Quaternary International 349 (2014) 257e269

Fig. 3. High-resolution d13C and d18O records of modern coral and environmental data. A. d13C and d18O of modern coral QG5; B. Solar radiation at sea surface (20 N, 110.625 E), and cloud cover at Qionghai meteorological station; C. Atmospheric CO2 concentration and d13C at Mauna Loa Observatory; D. SSS at Qinglan oceanic station, and rainfall at Qionghai meteorological station.

4.1.2. Seasonal variability in coral d13C and reconstruction of solar radiation in middle Holocene The annual d13C maxima and minima of the two middle Holocene corals consistently occur in AprileJune and NovembereDecember (Fig. 2AeB), respectively, which were similar to the modern coral's (Fig. 3A). This implies that strong seasonal offsets might happen between d13C and d18O of the two fossil corals. It suggests that the “kinetic” offset from equilibrium for both 18O and 13C appeared to have happened in the middle Holocene corals. The ~5.4 ka coral demonstrated a high value of d13C (Table 1). The average d13C was 18% higher than the modern's. The maximum

was 42% higher, more than the 3% increase of the d18O minimum mainly influenced by rainfall in summer; the minimum was 9% higher, less than the 15% increase of the d18O maximum, influenced by SSS controlled by solar radiation strengthening Asian monsoon in winter (Su et al., 2010). Moreover, the d13C range was 18% larger, which was smaller than the range increases 26% of the d18O and 25% of the radiation (Su et al., 2010), and 35% smaller than the range 4.6‰ of the ca. 4e7 ka coral from Belize (Gischler and Storz, 2009). These data show similar seasonal distributing characteristics but different variational amplitudes not only between fossil and modern coral d13C in the same region, but also between different regional corals in similar times. According to the regression

R. Su et al. / Quaternary International 349 (2014) 257e269

263

Fig. 4. High resolution detrending d13C of modern coral, solar radiation and linear regression result. A. Modern coral d13C and solar radiation; B. Regression analysis between the modern coral d13C and solar radiation. The regression equation is indicated correspondingly in the subdivision of the figure.

equation between the modern coral d13C and radiation (Fig. 4B), the reconstructed radiation (Fig. 5B) would be 194.02 W/m2 from the ~5.4 ka coral d13C (Fig. 5A), close to the calculated radiation of 193 W/m2 at ~5.4 ka BP (Su et al., 2010). This implies that radiation was well reconstructed by coral d13C at ~5.4 ka BP. Thus, higher values and increased seasonality of coral d13C might result from increasing radiation through strengthening photosynthesis, in the context of kinetic and metabolic effects, at ~5.4 ka BP. By contrast, the average and maximum of the ~4.4 ka coral d13C was 2% and 52% higher respectively than the modern coral's, but the minimum was 14% lower (Table 1). Comparatively, the maximal increase was more than the 8% increase of the d18O minimum mainly influenced by monsoonal rainfall in summer, whereas the minimal d13C decrease was opposite to the 14% increase of the d18O maximum, mainly influenced by SSS in winter (Sun et al., 2005). The seasonal amplitude increase of the ~4.4 ka coral d13C was 68% higher (Table 1), which was more than the increases of 12% of the d18O and 10% of the radiation at ~4.4 ka BP (Sun et al., 2005). According to the regression equation between the modern coral d13C and radiation (Fig. 4B), the reconstructed radiation would be 180.57 W/m2 from the ~4.4 ka coral d13C (Fig. 5C and D), which was close to the calculated radiation 171 W/m2 at ~4.4 ka BP (Sun et al., 2005), indicating that the ~4.4 ka coral d13C reconstructed radiation well. Thus, solar radiation is an important factor resulting in higher value and increased seasonality of coral d13C at ~4.4 ka BP.

However, coral d13C cycles peaked in late spring and then decreased during summer, somewhat ahead of seasonal light cycles, which suggests that coral d13C deviation from radiation might happen at any time, although it corresponded well to radiation in most of a year (Fig. 4A), implying that some other factors might affect seasonal variation in the fossil coral d13C. It is possible that nutrient level in water, the photosynthetic competence of zooxanthellae, and algal densities might be important (Franklin et al., 2006). Greater abundance of dissolved nutrient and planktonic food for coral probably contributed to high value and spring peak of coral d13C. Algal populations were probably higher in spring than in autumn, as zooxanthellae often show photosynthetic stress in spring and characteristically disperse during the summer, so that conditions might be more conducive to photosynthesis in winter and spring than in summer and autumn (Sun et al., 2008). Moreover, sea surface water generally becomes nutrient depleted after the midyear due to algal nutrient uptake and thermal stratification of the water column (Barton and Casey, 2005). Algal photosynthesis might remain depressed for much of the summer and autumn, and is restored to spring levels during wintertime nutrient replenishment and regrowth of zooxanthellae (Sun et al., 2008). Under higher concentrations of dissolved CO2 due to increase for transport efficiency of higher light (Geissler et al., 2009), symbiont algal photosynthesis and assimilation rate of photosynthetic CO2 might increase, so that host coral calcification was speeded up, and nutrients enriched. As a result,

264

R. Su et al. / Quaternary International 349 (2014) 257e269

Fig. 5. Reconstructed solar radiations from high resolution detrending data of two middle Holocene coral d13C. A. ~5.4 ka coral d13C; B. Reconstructed solar radiation with abnormity at ~5.4 ka BP; C. ~4.4 ka coral d13C; D. Reconstructed solar radiation with abnormity at ~4.4 ka BP. The reconstructed solar radiation is indicated by the solid line, and its abnormity is indicated by the broken line.

coral skeleton d13C increased (McConnaughey, 1989a, b). Overall, higher d13C values and seasonalities of both the ~5.4 and ~4.4 corals might be caused by higher light inducing stronger photosynthesis, and abundant nutrients. 4.2. Interannual and interdecade variability in coral d13C and atmospheric CO2 concentration and SSS in the middle Holocene 4.2.1. Interannual and interdecade variability in modern coral d13C and atmospheric CO2 concentration and SSS The annual value of the modern coral d13C shows a significant interannual variation (Fig. 6A). Meanwhile, it also shows a longterm decreasing trend at a rate of 1.71% decrease y 1 over the period of 1987e2002, which is opposite to the long-term increasing

trend in atmospheric CO2 at a rate of 0.41% increase y 1 over the same period (Fig. 6B). Linear regression analysis indicates a negative correlation between the coral d13C and atmospheric CO2 (Fig. 6C). According to the regression equation (Fig. 6C), the estimated atmospheric CO2 is 359.87 ppm from the modern coral d13C, being closer to 359.87 ppm for the instrumental atmospheric CO2. This shows a possible association between coral d13C and atmospheric CO2. The association may be illustrated through several processes or mechanisms. Firstly, increase in atmospheric CO2 due to anthropogenic CO2 increase has induced long-term decreases in atmospheric and seawater d13C and pH, and changes of relative proportions in DIC species (Fig. 3C; Keeling et al., 2009), which could decrease coral d13C (Nozaki et al., 1978; Su et al., 2007; Swart

R. Su et al. / Quaternary International 349 (2014) 257e269

265

Fig. 6. Annual data of modern coral d13C, atmospheric CO2 concentration and SSS and linear regression analysis. A. Modern coral QG5 d13C; B. ML atmospheric CO2 concentration and Qinglan SSS; C. Linear regression analysis between coral d13C and ML CO2; D. Linear regression analysis between coral d13C and Qinglan SSS. The trend and regression equations are indicated correspondingly in the subdivisions of the figure.

et al., 2010). Secondly, the addition of anthropogenically derived CO2 to the atmosphere has lead to climate warming, which could generally decrease coral d13C because coral respiration intensifies (McConnaughey, 1989a, b; Sun et al., 2008). Thirdly, nutrient level in surface seawater has been decreasing because of progressively intensified human activities, such as over fishing, environmental contamination, and physical destruction of the coral ecosystem from rapid economic growth in China (Sun et al., 2008). Thus, coral d13C decrease may partly be induced by atmospheric CO2 increase. However, the decreasing rate of the coral d13C is much larger than the increasing rate of the atmospheric CO2 (Fig. 6AeB), suggesting that coral d13C decrease might be induced not only by increasing atmospheric CO2 but also by other factors. Freshwater inputs display a negative relationship with both seawater SSS and coral d18O, which indicates that seawater SSS and oxygen isotope compositions are greatly affected by terrestrial runoff and rainfall (Sun et al., 2005; Su et al., 2006, 2010; Craig et al., 2010). Consistent variations may also occur in other seawater compositions. Carbon isotope compositions and microorganism compositions are important proxy indexes of such compositions. They are largely regulated by monsoon driving sea surface current through controlling terrestrial runoff and rainfall with lower d13C (Su et al., 2007). As a result, terrestrial biomass and ecosystems influence the abundance, composition, and timing of carbon delivered to ocean, particularly in shallow seas such as the northern

SCS, where coral reefs occur in proximity to the mouths of rivers. Thus, coral d13C may record d13C deviations in seawater DIC from SSS, indicating terrestrial runoff and rainfall (Swart et al., 1996; Su et al., 2007; Craig et al., 2010). Consistent decreasing trends between the modern coral d13C (Fig. 6A) and SSS (Fig. 6B) may confirm this interpretation. Regression analysis also reveals a significant positive relationship between them (Fig. 6D). According to the regression equation (Fig. 6D), the estimated SSS is 25.82‰, equivalent to the instrumental SSS 25.82‰. This shows that the two types of the SSSs correspond, so that SSS could be estimated from coral d13C. Thus, coral skeletal d13C provides a record of terrestrial carbon delivery to the coastal sea and coral reefs, as well as land-use history of the river catchment (Craig et al., 2010). Overall, decrease in modern coral d13C has been caused by increase in atmospheric CO2 and warming due to increasing anthropogenic CO2 emissions, progressive nutrient starvation, and SSS decrease indicating terrestrial runoff and rainfall increase in the northern SCS (Su et al., 2007; Sun et al., 2008). 4.2.2. Interannual and interdecade variability in coral d13C and atmospheric CO2 concentration and SSS in middle Holocene 4.2.2.1. Coral d13C and atmospheric CO2 concentration and SSS at ~5.4 ka BP. The d13C of the ~5.4 ka coral was 0.56‰ higher than the modern's (Figs. 6 and 7A; Table 1), indicating that atmospheric CO2 might be lower but SSS and nutrient level might be higher. The d13C

266

R. Su et al. / Quaternary International 349 (2014) 257e269

showed a long-term increasing trend (Fig. 7A), which was opposite to a long-term decreasing trend of the modern's (Fig. 6A), showing that atmospheric CO2 might decrease, but SSS and nutrient level might increase at ~5.4 ka BP. Increasing d13C of the ~5.4 ka coral was consistent with that of the natural atmospheric CO2 decrease in the Holocene, which was driven by orbitally forced summer insolation inducing ice volume increase (Milankovitch, 1941). It was probably the result of carbon uptake by the land biosphere and carbon release from the ocean in response to carbonate compensation of the terrestrial uptake (Joachim et al., 2009). Due to rapid air-sea gas exchange, the decrease in atmospheric CO2 caused a corresponding decrease in seawater CO2 concentration, and led to pH and carbonate

concentration increase (Furla et al., 2000; Bates, 2007), which promoted algal photosynthesis and host coral growth (Sabine et al., 2004). In this long linkage, coral d13C might follow seawater d13C increase through several mechanisms (Adkins et al., 2003). Firstly, as the metabolic demand for carbon increased with coral growth, the pool of available 12CO2 in the coral became relatively depleted, causing an increase to the d13C signature of coral (Erez, 1978; McConnaughey, 1989b). Secondly, following symbiont photosynthesis strengthening, high photosynthesis preferentially used 12C, leaving the DIC pool enriched in 13C (Goreau, 1977; Swart, 1983). If the correlation were significant, the estimated atmospheric CO2 would be 349.99 ppm from the regression equation between the modern coral d13C and

Fig. 7. Estimated atmospheric CO2 concentration and SSS from middle Holocene coral annual d13C. A. ~5.4 ka coral d13C; B. Estimated atmospheric CO2 concentration and SSS at ~5.4 ka BP; C. ~4.4 ka coral d13C; D. Estimated atmospheric CO2 concentration and SSS at ~4.4 ka BP. The trend equations are indicated correspondingly in the subdivisions of the figure.

R. Su et al. / Quaternary International 349 (2014) 257e269

atmospheric CO2 (Figs. 6C and 7A-B). Similarly, the 0.56‰ increase of the ~5.4 ka coral d13C relative to the modern coral's would require 9.88 ppm atmospheric CO2 decrease, while the 0.35‰ d13C increase over a period of 40 years would require 6.11 ppm atmospheric CO2 decrease. However, some estimations indicate a 0.2‰ d13C decrease induced by a 25 ppm atmospheric CO2 increase over a 7 ky period from 8 to 1 ka (Indermühle et al., 1999). Model simulations show that the total terrestrial carbon release from 6 to 0 ka BP would produce about a 5e7 ppm atmospheric CO2 increase (Joos et al., 2004; Wang et al., 2005). These data suggest that not only is our result opposite to the studied results of others, but also our estimated atmospheric CO2 decrease was much larger. It was impossible that all the 0.56‰ and 0.35‰ increases of the coral d13C were induced by higher and increasing d13C from atmospheric CO2 decrease at ~5.4 ka BP, which implies that other factors might be involved in higher and increasing coral d13C. Increasing salinity is a predominant stressor that affects survival, metabolism, and biomineralization response of calcifying organism (Matoo, 2014). It might increase CO2 gas exchange (Urban, 2003), algal photosynthetic rate, host calcification, and nutrient concentrations (Koyro, 2003). Especially, elevated CO2 utilization might lead to strengthen photosynthesis and increase photosynthetic efficiency of halophile organisms under saline conditions. Thus, host coral metabolism strengthened and the calcification increased following nutrient level increase (Geissler et al., 2009). As a result, coral d13C might increase, because strengthening photosynthesis preferentially used 12C, leaving the DIC pool enriched in 13C (Goreau, 1977; Swart, 1983). According to the regression equation between the modern coral d13C and SSS (Fig. 6D), the 0.56‰ increase of the coral d13C relative to the modern's would be induced by ~4.45‰ increase ofDSSS (Fig. 7AeB). Similarly, its maximal and minimal d13C were 0.87‰ and 0.42‰ higher (Table 1), respectively, which would be induced by ~6.91‰ and ~3.34‰ increase of maximal and minimalDSSS. The 0.35‰ increase over a period of 55 years would be induced by 2.75‰ SSS increase. However, reconstructed paleoSSSs showed increases with deviations of 0.5e2‰ during the middle Holocene relative to today in the SCS and northwestern Pacific (Stott et al., 2004; Sun et al., 2005; Morimoto et al., 2007; Su et al., 2010). Accordingly, higher and increasing d13C in the ~5.4 ka coral might be partly induced by higher and increasing SSS. Remarkable deviations of coral d13C from atmospheric CO2 and SSS might occur especially at its maximum in summer and minimum in winter. These deviations might be induced by other factors. High nutrient level in seawater was important for coral growth and 13C enrichment in coral skeleton (Barton and Casey, 2005). Climate might have been slightly warmer and more optimal in the northern SCS at ~5.4 ka BP (Su et al., 2010), and marine contamination might have been limited due to lesser physical destruction resulting from differences in energy use and socio-economic conditions in China. Moreover, the strengthened Asian monsoon might promote upper and under seawater mixture, nutritive material circulation at the sea bottom, and development of surface marine phytoplankton (Su et al., 2010), which could carry a large number of nutrients and dust containing many nutritive elements. These favorable conditions could promote increases of marine productivity, and zooxanthellae photosynthesis and coral growth (Franklin et al., 2006). As a result, coral d13C might be increased (Erez, 1978; Swart, 1983; McConnaughey, 1989b). Overall, higher and long-term increasing d13C of the ~5.4 ka coral was the result of higher and increasing 13C effect from lower and long-term natural decreasing atmospheric CO2, and higher and long-term increasing SSS and nutrient level.

267

Comparatively, the ss of the annual average, minimal and maximal d13C of the ~5.4 ka coral decreased 30% and 2% lower, but increased 3% higher relative to that of the modern coral (Table 1), which showed a weakened interannual change in coral d13C at ~5.4 ka BP. It is suggests that interannual variations of atmospheric CO2 concentration, SSS, and nutrient level influencing on the ~5.4 ka coral d13C might weaken. 4.2.2.2. Coral d13C and atmospheric CO2 concentration and SSS at ~4.4 ka BP. Similar to the ~5.4 ka coral, if the 0.06‰ increase of the ~4.4 ka coral d13C were produced by atmospheric CO2, the estimated atmospheric CO2 would be 358.86 ppm according to the regression equation between the modern coral d13C and atmospheric CO2 (Figs. 6C and 7C-D), lower than today. Similarly, if the coral d13C increase was the result of SSS change, the estimated SSS should increased by ~0.56‰ according to the regression equation between the modern coral d13C and SSS (Fig. 6D). The maximum and minimum of the coral d13C were 1.07‰ higher and 0.63‰ lower respectively, which would be resulted from ~8.50‰ minimal DSSS increase in summer and ~5.01‰ maximal DSSS decrease in winter (Fig. 7D). However, the estimated salinity minimal increase and maximal decrease were much larger than the reconstructed paleoSSSs with increasing deviations of 0.5e2‰ and decreasing deviations of 0e0.5‰ (Stott et al., 2004; Sun et al., 2005; Morimoto et al., 2007; Su et al., 2010). This suggests that remarkable deviations of coral d13C from atmospheric CO2 and SSS might happen especially at its maximum in summer and minimum in winter at ~4.4 ka BP, which might be induced by other factors. Higher d13C of the ~4.4 ka coral, similar to the ~5.4 ka coral's, might result from higher nutrient level due to warmer and stabilized climate, lower atmospheric seawater CO2, increasing pH and carbonate concentration, higher salinity, and strengthened East Asian monsoon. A combination of these reasons could promote algal photosynthesis strengthening and increase nutrient level. As a result, symbiont-host metabolism might intensify, and coral d13C might increase following d13C increase in calcifying fluid. Thus, higher d13C in the ~4.4 ka coral were influenced not only by lower atmospheric CO2 and higher SSS, but also by higher nutrient level. The annual d13C of the ~4.4 ka coral showed a decreasing trend with 0.40‰ decrease at a rate of 0.26 decrease y 1 over a period of 55 years (Fig. 7C), which is consistent with the modern coral's (Fig. 6A), but opposite to the ~5.4 ka coral's (Fig. 7A). This trend might be the combined consequence of long-term decreasing d13C in seawater from increasing atmospheric CO2, and decreasing SSS and nutrient level. Paleo-CO2 studies suggest that, theoretically, atmospheric CO2 should naturally decrease during the Holocene (Milankovitch, 1941). However, it began to deviate from its natural decreasing trend some 8 ka ago, so that it prevented the climate system from entering a new glacial (Brovkin et al., 2007; Ruddiman et al., 2011). Perhaps, an unusual increase in atmospheric CO2 might have happened much later, which might be consistent with the unusual increase in atmospheric CH4 at ~5.2 ka BP (Blunier et al., 1994). Thus, it was possible that atmospheric CO2 unusually increased at the SCS at ~4.4 ka BP. Possible unusual increase in atmospheric CO2 in the northern SCS at ~4.4 ka BP could be explained by contributions from decrease in carbonate compensation of earlier land-biosphere uptake (Elsig et al., 2009; Joachim et al., 2009), coral reef formation (Ridgwell et al., 2003) and terrestrial carbon storage (Indermühle et al., 1999; Brovkin et al., 2007), and increase in SST (Monnin et al., 2001; Joos et al., 2004). Plants prefer 12 C over 13C, so that d13C in plants is less than in the atmosphere. As CO2 originally came from plant metabolism, more 12C was released into the atmosphere. If CO2 releases were responsible for the rise in atmospheric CO2 levels, the d13C should be decreased (Swart et al., 2010; Dassie et al., 2013). Atmospheric d13C should decrease as

268

R. Su et al. / Quaternary International 349 (2014) 257e269

increasing plant metabolism, especially respiration, and anthropogenic emissions of CO2 increased the levels of atmospheric CO2 concentration. Based on the relationship between the modern coral and atmosphere CO2 (Fig. 6C), the 0.40‰ decrease of the ~4.4 ka coral d13C over a period of 55 years would require 7.12 ppm atm CO2 increase (Fig. 7CeD). However, studied results proposed that the 25 ppm atm CO2 increase over a 7.2-ka period that induced the 0.2‰ atmospheric d13C decrease (Indermühle et al., 1999); Anthropogenic land use might have caused 40 ppm CO2 increase and 0.6‰ d13C decrease during the Holocene (Ruddiman et al., 2011); Model simulations show that the total terrestrial carbon release from 6 to 0 ka BP would produce about a 5e7 ppm atmospheric CO2 increase (Joos et al., 2004). Thus, it was impossible that all the 0.40‰ decrease was induced by atmospheric CO2 increase. According to the regression equation between the modern coral d13C and SSS (Fig. 6D), the 0.40‰ coral d13C decrease would require 3.21‰ SSS decrease (Fig. 7CeD). This value is much larger than the reconstructed paleo-SSS decreases with deviations of 0e0.5‰ (Stott et al., 2004; Sun et al., 2005; Morimoto et al., 2007; Su et al., 2010). Thus, remarkable deviations of coral d13C from atmospheric CO2 and SSS might occur at ~4.4 ka BP (Sun et al., 2008). It implies that other factors might play a role in coral d13C trend variation, such as nutrient level. Rise in atmospheric CO2 caused a corresponding increase in seawater CO2 concentrations in upper ocean waters due to rapid air-sea gas exchange, which resulted in pH and carbonate concentration decrease (Furla et al., 2000; Bates, 2007). Following assimilation rate of algal photosynthetic CO2 increase under SSS and pH decrease, many aspects of symbiont algae growth might be led to be weakened, such as, photosynthesis and productivity, internal chemical composition, physiological and biochemical mechanisms, so that host coral nutrient chain affected by symbiont algae might be impacted. As a result, symbiont photosynthesis weakened and coral calcification decreased but coral respiration increased, providing decreasing d13C to coral calcifying fluid of DIC pool, following which coral d13C might decrease. Long-term decreasing trend in coral d13C might be induced by long-term decreasing d13C due to unusual increasing atmospheric concentration, and decreasing SSS and nutrient level at ~4.4 ka BP. In addition, the ss of the ~4.4 ka coral average, maximal and minimal d13C consistently increased 27%, 103% and 10% higher respectively relative to that of the modern coral (Table 1), indicating that interannual variability of coral d13C might be intensified especially in summer at ~4.4 ka BP, which is consistent with that of its d18O related to the intensified Asian monsoon that would directly correlate with orbitally strengthened interannual variability in solar radiation in the middle Holocene (Sun et al., 2005; Su et al., 2010). Therefore, intensified interannual variation of the ~4.4 ka coral d13C might imply intensifying interannual changes of atmosphere CO2 concentration, SSS indicating seawater components related to terrestrial runoff and rainfall, and nutrient level under the strengthened Asian monsoon, especially in summer. 5. Conclusions 1. Seasonal change of the middle Holocene coral d13C was in phase with solar radiation and its most seasonal features were modulated by light and nutrient level inducing photosynthesis. Thus, higher value and seasonality of coral d13C was ascribed to light inducing stronger photosynthesis. Higher coral d13C was also the result of higher d13C in atmospheric and seawater due to lower atmospheric and seawater CO2 concentration, higher SSS and abundant nutrients due to less anthropogenic activities and strengthened Asian monsoon under optimum climate conditions during the middle Holocene.

2. Weakened interannual variation of the ~5.4 ka coral d13C might be related to weakened interannual changes of atmosphere and seawater CO2 concentration, SSS, and nutrient level. By contrast, intensified interannual variation of the ~4.4 ka coral d13C might be induced by intensified interannual changes of atmosphere and seawater CO2 concentration, SSS indicating seawater components related to terrestrial runoff and rainfall, and nutrient level at ~4.4 ka BP. 3. A long-term increasing trend in the ~5.4 ka coral d13C was attributed to increasing atmospheric and seawater d13C due to a long-term natural decrease in atmospheric CO2 concentration driven by orbitally forced summer insolation inducing ice volume increase, and increasing SSS and nutrient level. However, a long-term decreasing trend in the ~4.4 ka coral d13C might be caused by decreasing atmospheric and seawater d13C due to unusual increasing atmospheric CO2 concentration, and decreasing SSS and nutrient level.

Acknowledgments We cordially thank Norm Catto, Editor-in-Chief of Quaternary International, for earnestly taking charge of the manuscript review process; reviewers, for strictly and fairly commenting on our manuscript and making many useful suggestions. Financial support for this research was provided by the National Science Foundation of China (Grant 41176051, 41272045, 40876025 and 40625009).

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quaint.2014.08.030.

References Adkins, J.F., Boyle, E.A., Curry, W.B., 2003. Stable isotopes in deep-sea corals and a new mechanism for ‘‘vital effects’’. Geochimica et Cosmochimica Acta 67, 1129e1143. Barton, A.D., Casey, K.S., 2005. Climatological context for largescale bleaching. Coral Reefs 24, 536e554. Bates, N.R., 2007. Interannual variability of the oceanic CO2 sink in the subtropical gyre of the North Atlantic Ocean over the last 2 decades. Journal of Geophysical Research-Atmospheres 112 (C09013), 1e26. Blunier, T., Chappellaz, J., Schwander, J., Stauffer, B., Raynaud, D., 1994. Variations in atmospheric methane concentration during the Holocene epoch. Nature 374, 46e49. Brovkin, V., Ganopolski, A., Archer, D., Rahmstorf, S., 2007. Lowering of glacial atmospheric CO2 in response to changes in oceanic circulation and marine biogeochemistry. Paleoceanography 22, PA4202. http://dx.doi.org/10.1029/ 2006PA001380. Coles, S.L., Jokiel, P.L., 1978. Synergistic effects of temperature, salinity and light on the hermatypic coral Montipora verrucosa. Marine Biology 49, 187e195. Craig, G., Tim, S., Jens, Z., Eric, E., Wim, B., Bemahafaly, R., Geert-Jan, B., 2010. Linkages between coral proxies and terrestrial runoff across largest Madagascar embayment. International Society for Reef Studies. Abstract book. http:// togather.eu/handle/123456789/1004. Dassie, E.P., Lemley, G.M., Linsley, B.K., 2013. The Suess effect in Fiji coral d13C and its potential as a tracer of anthropogenic CO2 uptake. Palaeogeography, Palaeoclimatology, Palaeoecology 370, 30e40. Elsig, J., Schmitt, J., Leuenberger, D., Schneider, R., Eyer, M., Leuenberger, M., Joos, F., Fischer, H., Stocker, T.F., 2009. Stable isotope constraints on Holocene carbon cycle changes from an Antarctic ice core. Nature 461, 507e510. Erez, J., 1978. Vital effect on stable-isotope composition seen in foraminifera and coral skeletons. Nature 273, 199e202. Fairbanks, R.G., Dodge, R.E., 1979. Annual periodicity of 18O/16O and 13C/12C ratios in the coral Montastrea annularis. Geochimica et Cosmochimica Acta 43, 1009e1020. s, C., Hoegh-Guldberg, O., 2006. Increased mortality and phoFranklin, D.J., Cedre toinhibition in the symbiotic dinoflagellates of the IndoePacific coral Stylophora pistillata (Esper) after summer bleaching. Marine Biology 149, 633e642. Furla, P., Galgani, I., Durand, I., Allemand, D., 2000. Sources and mechanisms of inorganic carbon transport for coral calcification and photosynthesis. Journal of Experimental Biology 203, 3445e3457.

R. Su et al. / Quaternary International 349 (2014) 257e269 Gagan, M.K., Chivas, A.R., Isdale, P.J., 1996. Timing coral-based climatic histories using 13C enrichments driven by synchronized spawning. Geology 24, 1009e1012. Gischler, E., Storz, D., 2009. High-resolution windows into Holocene climate using proxy data from Belize corals (Central America). Palaeobiodiversity and Palaeoenvironments 89, 211e221. Geissler, N., Hussin, S., Koyro, H.-W., 2009. Elevated atmospheric CO2 concentration ameliorates effects of NaCl salinity on photosynthesis and leaf structure of Aster tripolium L. Journal of Experimental Botany 60, 137e151. Goreau, T.J., 1977. Carbon metabolism in calcifying and photosynthetic organisms: theoritical models based on stable isotope data. In: Third International Coral Reef Symposium, pp. 395e401. Grottoli, A.G., Rodrigues, L.J., Juarez, C., 2004. Lipids and stable carbon isotopes in two species of Hawaiian corals, Porites compressa and Montipora verrucosa, following a bleaching event. Marine Biology 145, 621e631. Grottoli, A.G., Wellington, G.M., 1999. Effect of light and zooplankton on skeletal d13C values in the eastern Pacific corals Pavona clavus and Pavona gigantea. Coral Reefs 18 (1), 29e41. Grove, C.A., Zinke, J., Scheufen, T., Maina, J., Epping, E., et al., 2012. Spatial linkages between coral proxies of terrestrial runoff across a large embayment in Madagascar. Biogeosciences 9.8, 3063e3081. Gupta, L.P., Suzuki, A., Kawahata, H., 2006. Aspartic acid concentrations in coral skeletons as recorders of past disturbances of metabolic rates. Coral Reefs 25, 599e606. Hemming, N.G., Guilderson, T.P., Fairbanks, R.G., 1998. Seasonal variances in the boron isotopic composition of coral: a productivity signal? Global Biogeochemical Cycles 12, 581e586. Indermühle, A., Stocker, T.F., Joos, F., Fischer, H., Smith, H.J., Wahlen, M., Deck, B., Mastroianni, D., Tschumi, J., Blunier, T., Meyer, R., Stauffer, B., 1999. Holocene carbon-cycle dynamics based on CO2 trapped in ice at Taylor Dome, Antarctica. Nature 398, 121e126. Joachim, E., Jochen, S., Daiana, L., Robert, S., Marc, E., et al., 2009. Stable isotope constraints on Holocene carbon cycle changes from an Antarctic ice core. Nature 461, 507e510. Joos, F., Gerber, S., Prentice, I.C., Otto-Bliesner, B.L., Valdes, P.J., 2004. Transient simulations of Holocene atmospheric carbon dioxide and terrestrial carbon since the Last Glacial Maximum. Global Biogeochemical Cycles 18, GB2002. http://dx.doi.org/10.1029/2003GB002156. Keeling, R.F., Piper, S.C., Bollenbacher, A.F., Walker, S.J., 2009. Atmospheric CO2 Concentrations (Ppmv) Derived from in Situ Air Samples Collected at Mauna Loa Observatory, Hawaii, USA. Carbon Dioxide Research Group, Scripps Institution of Oceanography (SIO), University of California, La Jolla, California, USA 92093e0444. Keith, M.L., Weber, J.N., 1965. Systematic relationships between carbon and oxygen isotopes in carbonates deposited by modern corals and algae. Science 150, 498e501. Koyro, H.-W., 2003. Study of potential cash crop halophytes by a quick check system: determination of the threshold of salinity tolerance and the ecophysiological demands. In: Lieth, H., Mochtchenko, M. (Eds.), Cash Crop Halophytes: Recent Studies, Tasks for Vegetation Science, vol. 38. Kluwer Academic Publishers, Dordrecht, the Netherlands, pp. 5e17. Levy, O., Rosenfeld, M., Loya, Y., Yam, R., Mizrachi, I., et al., 2010. Anthropogenic stressors and eutrophication processes as recorded by stable isotopes compositions in coral skeletons. Biogeosciences Discussions 7, 7657e7672. Levy, O., Dubinsky, Z., Achituv, Y., Erez, J., 2006. Diurnal polyp expansion behavior in stony corals may enhance carbon availability for symbionts photosynthesis. Journal of Experimental Marine Biology and Ecology 333, 1e11. McConnaughey, T., 1989a. 13C and 18O isotopic disequilibrium in biological carbonates: I. Patterns. Geochimica et Cosmochimica Acta 53, 151e162. McConnaughey, T., 1989b. 13C and 18O isotopic disequilibrium in biological carbonates: II. In vitro simulation of kinetic isotope effects. Geochimica et Cosmochimica Acta 53, 163e171. Matoo, O.B., 2014. Interactive Effects of Ocean Acidification and Multiple Stressors on the Physiology of Marine Bivalves. University of North Carolina, p. 3608352 (PhD, UMI). Milankovitch, 1941. Kanon der Erdbestrahlung und seine Andwendung auf das Eiszeitenproblem, vol. 133. Royal Serbian Academy Special Publication, Belgrade, p. 633 (English translation published in 1969 by Israel Program for Scientific Translations, US Dept. Comm.). Monnin, E., Indermühle, A., D€ allenbach, A., Flückiger, J., Stauffer, B., Stocker, T.F., Raynaud, D., Barnola, J.M., 2001. Atmospheric CO2 concentrations over the last glacial Termination. Science 291, 112e114.

269

Morimoto, M., Kayanne, H., Abea, O., McCullochc, M.T., 2007. Intensified midHolocene Asian monsoon recorded in corals from Kikai Island, subtropical NP. Quaternary Research 67, 204e214. Nozaki, Y., Rye, D.M., Turekian, K.K., Dodge, R.E., 1978. A 200 year record of carbon13 and carbon-14 variations in a Aermuda coral. Geophysical Research Letters 5, 825e828. s, C., Gattuso, J.-P., Kamber, B.S., Samankassou, E., Pretet, C., Reynaud, S., Ferrier-Page 2014. Effect of salinity on the skeletal chemistry of cultured scleractinian zooxanthellate corals: Cd/Ca ratio as a potential proxy for salinity reconstruction. Coral Reefs 33, 169e180. Quinn, T.M., Taylor, F.W., Crowley, T.J., 1993. A 173 year stable isotope record from a tropical south Pacific coral. Quaternary Science Reviews 12, 407e418. Ridgwell, A.J., Watson, A.J., Maslin, M.A., Kaplan, J.O., 2003. Implications of coral reef buildup for the controls on atmospheric CO2 since the Last Glacial Maximum. Paleoceanography 18 (4), 1083. http://dx.doi.org/10.1029/2003PA000893. Rollion-Bard, C., Blamart, D., Cuif, J.-P., Juillet-Leclerc, A., 2003. Microanalysis of C and O isotopes of azooxanthellate and zooxanthellate corals by ion microprobe. Coral Reefs 22, 405e415. Ruddiman, W.F., Kutzbach, J.E., Vavrus, S.J., 2011. Can natural or anthropogenic explanations of late-Holocene CO2 and CH4 increases be falsified? The Holocene. http://dx.doi.org/10.1177/0959683610387172. Sabine, C.L., Feely, R.A., Gruber, N., Key, R.M., Lee, K., Bullister, J.L., Wanninkhof, R., Wong, C.S., Wallace, D.W.R., Tilbrook, B., Millero, F.J., Peng, T.-H., Kozyr, A., Ono, T., Rios, A.F., 2004. The oceanic sink for anthropogenic CO2. Science 305, 367e371. Smith, S.V., Kroopnick, P., 1981. Carbon-13 isotopic fractionation as a measure of aquatic metabolism. Nature 294, 252e253. Stott, L., Cannariato, K., Thunell, R., Haug, G.H., Koutavas, A., Lund, S., 2004. Decline of surface temperature and salinity in the west tropical Pacific Ocean in the Holocene epoch. Nature 431, 56e59. Su, R., Sun, D., Chen, H., Chen, X., Li, Z., 2010. Evolution of Asian monsoon variability revealed by oxygen isotopic record of middle Holocene massive coral in the northern South China Sea. Quaternary International 213, 56e68. Su, R., Sun, D., Zhu, Z., Chen, X., 2007. Temporal and spatial variation of carbon isotopic composition of coral Porites in the South China Sea and the environmental causes. Quaternary Sciences 27, 133e148 (in Chinese). Su, R., Sun, D., Bloemendal, J., Zhu, Z., 2006. Temporal and spatial variability of the oxygen isotopic composition of massive corals from the South China Sea: influence of the Asian monsoon. Palaeogeography Palaeoclimatology Palaeoecology 240, 630e648. Sun, D., Su, R., McConnaughey, T.A., Bloemendal, J., 2008. Skeletal growth and d13C in massive corals from South China Sea: effects of photosynthesis, respiration and human activities. Chemical Geology 255, 414e425. Sun, D., Gagan, M.K., Cheng, H., Scott-Gagan, H., Dykoski, C.A., Edwards, R.L., Su, R., 2005. Seasonal and interannual variability of the Mid-Holocene East Asian monsoon in coral d18O records from the South China Sea. Earth and Planetary Science Letters 237, 69e84. Swart, P.K., Greer, L., Rosenheim, B.E., Moses, C.S., Waite, A.J., Winter, A., Dodge, R.E., Helmle, K., 2010. The 13C Suess effect in scleractinian corals mirror changes in the anthropogenic CO2 inventory of the surface oceans. Geophysical Research Letters 37 (L05604), 1e5. http://dx.doi.org/10.1029/2009GL041397. Swart, P.K., Szmant, A., Porter, J.W., Dodge, R.E., Tougas, J.I., et al., 2005. Implications for cycling of organic carbon in corals. Geochimica et Cosmochimica Acta 69 (6), 1495e1509. Swart, P.K., Dodge, R.E., Hudson, H.J., 1996. A 240-year stable oxygen and carbon isotopic record in a coral from south Florida: implications for the prediction of precipitation in south Florida. Palaios 11, 362e375. Swart, P.K., 1983. Carbon and oxygen isotope fractionation in scleractinian corals: a review. Earth Science Reviews 19, 51e80. Urban, O., 2003. Physiological impacts of elevated CO2 concentration ranging from molecular to whole plant responses. Photosynthetica 41, 9e20. Wang, Y., Mysak, L.A., Roulet, N.T., 2005. Holocene climate and carbon cycle dynamics: experiments with the “green” McGill Paleoclimate Model. Global Biogeochemical Cycles 19 (3). GB3022/1eGB3022/18. Weber, J.N., Deines, P., Weber, P.H., Baker, P.A., 1976. Depth related changes in the 13 12 C/ C ratio in skeletal carbonate deposited by the Caribbean reef-frame building coral Monstatrea annularis: futher implications of a model for stable isotope fractionation by Scleractinian corals. Geochimica et Cosmochimica Acta 40, 31e39.