Diatoms from the Pearl River estuary, China and their suitability as water salinity indicators for coastal environments

Abstract

We collected 77 modern diatom samples from sites across the Pearl River estuary, China to analyze the relationship between diatom assemblages and environmental parameters including water salinity, water depth and sediment particle size. Results showed that marine diatoms were dominant in the high salinity environment around Hong Kong and the outer part of the estuary. Brackish water diatoms were found in high abundance in the central part of the estuary. Both marine and brackish water diatoms were predominantly planktonic taxa. Freshwater diatoms dominated in low salinity environments, with planktonic taxa in the deep tidal channel and benthic species in the shallow deltaic distributaries. Statistical tests indicated that the modern diatom distribution is strongly correlated with salinity but is also influenced by several other environmental variables including sand content and water depth. Transfer functions relating diatom assemblages and water salinity had high r² (0.94–0.98 for WA-PLS, 0.95 for MA and 0.98 for MAT) and relatively low RMSEP (2.66–1.63‰ for WA-PLS, 2.35‰ for MA and 2.70‰ for MAT). Due to the geographical distribution of samples, some spatial autocorrelation is likely present in the dataset. When this effect is considered, r² decreases to 0.90 and RMSEP increases to 5.41‰, although the diatom–salinity relationship remains appropriate for reconstructing paleosalinity. Based on this estimate, a diatom-based salinity transfer function with high accuracy and precision is developed and successfully applied to a sediment core for quantitative reconstruction of the Holocene paleosalinity in the Pearl River estuary.

Introduction

Variability in freshwater river discharge is an important factor that causes changes in the sedimentary regime, salinity and bio-productivity within estuarine environments (Chen et al., 2007; Zhang et al., 2008; Zong et al., 2009a). Since freshwater discharge is closely related to precipitation patterns within a given river catchment, changing freshwater flux can be used to infer changes in climate conditions. In the case of Asian Monsoon, long-term changes in summer precipitation drive fluctuations in the freshwater flux entering an estuary (Clift and Plumb, 2008). During stronger summer monsoons a greater amount of freshwater is discharged into an estuary, and vice versa (Zong et al., 2006). Holocene changes in summer monsoon strength have also resulted in modified terrestrial sediment supply, and directly impacted the rate of deltaic progradation (Zong et al., 2009a). Between 6800 and 2000 cal. years BP, for example, the rate of deltaic shoreline progradation decreased gradually as a result of a weakening summer monsoon (Zong et al., 2009a). Such reconstructions of temporal changes in paleo-monsoon strength are important for improving the predictive power of global and regional climate models. Given the recent problems with winter freshwater supply in the Pearl River region, accurate knowledge of variability in freshwater discharge in the recent past is critical for assisting local communities to plan better for their future freshwater consumption. Such knowledge will also be useful to societies living on the coasts of East Asia and beyond, because water supply is becoming an important issue for human society in the 21st century (Zhang et al., 2008).

Most recent studies regarding Holocene Asian summer monsoon strength are based on loess sequences (e.g. An et al., 2004), marine deposits for Sea Surface Temperature (SST; e.g. Wang et al., 2005), lake sequences for local productivity and vegetation history (e.g. Mingram et al., 2004, Shen et al., 2006), and stalagmites (e.g. Wang et al., 2008). Few attempts have been made to reconstruct the history of monsoonal driven freshwater flux. Zong et al. (2006) first explored the possibility of using diatoms as an indicator of freshwater flux in a monsoon region, and examined the qualitative relationship between diatom assemblages and water salinity. Zong et al. (2009a) further successfully applied the diatom technique to reconstruct the sedimentary history of the Pearl River delta. Elsewhere (in the UK, USA, Japan and Southeast Asia), diatoms have been widely used to reconstruct sea-level history (e.g. Long, 1992, Shennan et al., 1993, Shennan et al., 1994, Zong and Tooley, 1996, Zong, 1997b, Zong, 1998), ice sheet dynamics (e.g. Stabell, 1980, Long et al., 1999, Long et al., 2008), earthquake induced land movements (Shennan et al., 1996, Shennan et al., 1999, Long and Shennan, 1998, Sawai, 2001, Sawai, 2009) and coastal evolution (Juggins, 1992, Zong, 1992, Zong, 1997a, Zong, 1997b, Zong, 1998, Zong and Tooley, 1999). The success of these studies was based upon understanding the modern relationship between diatoms and coastal environments (e.g. Hendey, 1964, Sullivan, 1975, Sullivan, 1978, Denys, 1991/2, 1991/2, Vos and de Wolf, 1993, Nelson and Kashima, 1993, Hemphill-Haley, 1995, Zong, 1997a, Zong and Horton, 1998, Zong and Kamaludin, 2004, Sawai, 2001). Furthermore, reconstruction of coastal environmental change using fossil diatoms has become more precise and accurate when statistical methods such as transfer functions are used to establish the diatom-environment relationship quantitatively (e.g. Zong and Horton, 1999, Zong et al., 2003, Hamilton and Shennan, 2005, Horton et al., 2007).

In order to improve the reconstruction of monsoon-induced freshwater discharge using diatoms as explored by Zong et al. (2006), it is important to test the relationship between diatoms and a number of environmental factors including water salinity, water depth and the nature of substrate, as these factors influence population change and spatial distribution patterns of diatoms. For instance, planktonic diatoms distribute according to water salinity from tidal influenced river channels through estuarine environments to marine waters (Juggins, 1992, Zong et al., 2006), whilst benthic diatoms are sensitive to the combined effect of water salinity, substrate sediment type and ground altitude, the latter in the form of tidal inundation frequency (e.g. Nelson and Kashima, 1993, Hemphill-Haley, 1995, Zong and Horton, 1998, Zong and Kamaludin, 2004). The relationship between diatoms and water depth has not been examined, but preliminary results from Zong et al. (2006) suggested that the percentage of benthic diatoms increased toward the distributaries where water is shallower. Therefore, prior to reconstructing water salinity and monsoonal freshwater discharge in the Pearl River using diatom-based salinity transfer functions it is necessary to examine modern diatom distribution patterns and the relative influence of other environmental factors. In this paper we document the distributional patterns of modern diatoms across the Pearl River estuary and analyze the statistical relationship between diatom assemblages and a number of other appropriate environmental variables. We develop and test the applicability of the first diatom-based salinity transfer functions to aid reconstruction of freshwater flux history from estuarine sediments of the Pearl River estuary and beyond.