Seasonal changes in sediment erodibility in a sandy carbonate environment detected from turbidity time series

Highlights

• Standard deviation of turbidity can be used as a proxy for sand resuspension.

• Sand biostabilization was detected from continuous near-bed turbidity data.

• Annual cycle of biostabilization likely occurs in Hamelin Pool, Western Australia.

Abstract

Sediment resuspension and sediment erodibility were investigated in Hamelin Pool (Western Australia), which is one of the few modern carbonate settings where stromatolites and ooids are actively forming. For this purpose, the first time series of wave height and water turbidity were recorded at a ~2.5 m deep subtidal location in Hamelin Pool from July 2017 to March 2018. Sediment grain size analyses indicate that this environment is nearly free of mud, so the optical turbidity sensor was able to detect the intra-wave modulation of sand resuspension. This allowed us to use the standard deviation of the turbidity within a burst – as opposed to the mean turbidity – as a proxy for sand resuspension. This approach was further validated by wave tank experiments, which also helped to identify the occurrence of biofouling and discard the affected data. Although the wave regime was similar throughout the year, sand resuspension changed with seasons. Sediment erodibility, defined as the sand resuspension normalized by the bed shear stress, was higher during the Austral winter and spring, and lower in summer and autumn. This decrease in erodibility coincided with an increase in temperature and solar irradiance. These two environmental factors likely modulate the growth of benthic microbes, which in turn affects sediment dynamics. This seasonal modulation of sediment dynamics may influence additional processes, including the delivery of sand grains to stromatolites and the precipitation of aragonite around ooids and in stromatolite laminae.

Introduction

Benthic microbial growth is seasonal due to its dependence on light and temperature (MacIntyre et al., 1996; van der Wal et al., 2008; Bowlin et al., 2012; de Jonge et al., 2012; Benyoucef et al., 2014), availability of nutrients (Sundbäck and Snoeijs, 1991; Hillebrand and Sommer, 1997), and hydrodynamic disturbances such as tides and waves (Mariotti and Fagherazzi, 2012; Mariotti et al., 2014). Because benthic microbes can stabilize sediments (Yallop et al., 1994; Kornman and Deckere, 1998; Lucas et al., 2003; Le Hir et al., 2007) – a process known as biostabilization – seasonal variability in microbial growth could have an important role in modulating sedimentary processes.

The occurrence of seasonal biostabilization has been extensively studied in muddy environments by direct measurements of the critical bed shear stress. For example, some studies used the Cohesive Strength Meter (Friend et al., 2003; Tolhurst et al., 2008; De Backer et al., 2010; Waqas et al., 2020), others used the Gust Erosion Chamber (Wiberg et al., 2013; Chen et al., 2017). In temperate climates, mud biostabilization during warmer months results in low erodibility because higher temperatures favor microbial growth (Underwood and Paterson, 1993; Andersen, 2001; Wiberg et al., 2013; Schmidt et al., 2016). Yet, these field measurements provide a coarse temporal resolution and limit the ability to detect and quantify seasonal changes in biostabilization.

In muddy environment, detection of sediment resuspension from turbidity measurements is very unlikely due to practical constraints. Biostabilization decreases sediment resuspension and should lead to a lower water turbidity (Pivato et al., 2019). Therefore, measurements of turbidity (normalized by the amount of resuspension intensity, e.g., the bed shear stress) could be used to detect changes in sediment erodibility. Although possible in principle, this approach has not been successful for at least two reasons. First, turbidity time series spanning multiple seasons are difficult to obtain. Second, the detection of biostabilization through the analysis of turbidity time series is problematic in muddy environments. Indeed, not only does the amount of suspended mud depend on local resuspension, but it also on lateral advection (Hill et al., 2003). Hence, unless biostabilization occurs over very large areas, e.g., several kilometers or more, it might not be clearly reflected in the turbidity signal. Therefore, effort to relate turbidity data with mud resuspension is ineffective.

Because of its high settling velocity, sand is suspended and settles back within a single wave cycle (Trouw et al., 2000; Davies and Thorne, 2005; Williams and Bell, 2006; Murray et al., 2012), producing a strong modulation of the turbidity signal. Therefore, the use of turbidity as an indicator of biostabilization is more likely to be effective in sandy environments, given that sand responds nearly instantaneously to local hydrodynamics. In practice, however, siliciclastic sandy environments always have some small amounts of mud that dominate the turbidity signal because the optical backscatter is one order of magnitude more sensitive to mud than to sand (Conner and De Visser, 1992). In practice this limits the usefulness of turbidity as a proxy for sand dynamics. Acoustic measurements might be used to measure sand dynamics even in the presence of mud (Thorne and Hanes, 2002), but the deployment of acoustic sensors in the field is difficult over long periods and is generally shorter than few months (Christensen et al., 2019).

Carbonate-depositing environments often lack terrigenous mud, although they can occasionally contain carbonate mud (Bathurst, 1971). It is plausible that sand resuspension can be accurately detected via turbidity measurements and possibly used to identify biostabilization and track its seasonality in areas where mud is demonstrably absent. Here, this idea was tested by focusing on Hamelin Pool (Western Australia), a hypersaline coastal embayment that is demonstrably mud-poor. This site is of particular relevance because it hosts benthic microbes that enable the formation of stromatolites and other lithified microbial structures (Jahnert and Collins, 2012).

We were not aware of any previous reports of the wave characteristics in Hamelin Pool, so hourly near-bed pressure and turbidity data were recorded over nine months at one subtidal location at this site. Wave tank laboratory experiments were also used to develop a simple method to distinguish the contributions of sand and mud resuspension to the turbidity signal. Analyses of the time series from Hamelin Pool detected seasonal changes in sand erodibility and were used to infer the occurrence of biostabilization.