Jaco H Baas, Megan L. Baker, Jonathan Malarkey , Sarah J. Bass, Andrew J. Manning, Julie A. Hope, Jeff Peakall, Ian D. Lichtman, Leiping Ye, Alan G. Davies

The shape and size of sedimentary bedforms play a key role in the reconstruction of sedimentary processes in modern and ancient environments. Recent laboratory experiments have shown that bedforms in mixed sand–clay develop at a slower rate and often have smaller heights and lengths than equivalent bedforms in pure sand. This is generally attributed to cohesive forces that can be of physical origin, caused by electrostatic forces of attraction between clay minerals, and of biological origin, caused by ‘sticky’ extracellular polymeric substances (EPS) produced by micro-organisms, such as microalgae (microphytobenthos) and bacteria. In the present paper, we demonstrate, for the first time, that these laboratory experiments are a suitable analogue for current ripples formed by tidal currents on a natural mixed sand–mud–EPS intertidal flat in a macrotidal estuary. Moreover, both the field data and the laboratory data demonstrate that the widely used definitions of ‘clean sand’ (<25% mud: Shepard, 1954) and ‘mature sandstone’ (arenite, <10-15% mud: Folk, 1951; Dott, 1964) need to be redefined. Integrated hydrodynamic and bed morphological measurements, collected during a spring tide near Hilbre Island (Dee estuary, NW England), reveal a statistically significant linear decease in current ripple length for progressively higher bed mud contents, and a concurrent change from three-dimensional linguoid to two-dimensional straight-crested ripple plan morphology. These results agree well with observations in laboratory flumes, but the rate of decrease of ripple length was found to differ substantially between the field and the laboratory. Since the formation of ripples under natural conditions is inherently more complex than in the laboratory, five additional controls that might affect current ripple development in estuaries, but have not been accounted for in laboratory experiments, were explored: wave energy, flow energy, clay type, pore water salinity, and bed EPS content. This analysis showed that wave energy and clay type cannot be used to explain the difference in the rate of decrease in ripple length, because surface water waves were weak during the flood and ebb tides preceding the ripple length measurements, and the bed clay contents were too low for clay type to have had a measurable effect on bedform development. Accounting for the differences in flow forcing between the field and experiments, and therefore the relative stage of development with respect to equilibrium ripples, increases the difference between the ripple lengths by 50%. The presence of strongly cohesive EPS in the current ripples on the natural intertidal flat might explain most of the difference in the rate of decrease in ripple length between the field and the laboratory. The effect of pore water salinity on the rate of bedform development cannot be quantified at present, but salinity is postulated herein to have had a smaller influence on the ripple length than bed EPS content. The common presence of clay and EPS in many aqueous sedimentary environments implies that a re-assessment of the role of current ripples and their primary current lamination in predicting and reconstructing flow regimes is necessary, and that models that are valid for pure sand are an inappropriate descriptor for more complex mixed sediment. We propose that this re-assessment is necessary at all bed clay contents above 3%. This bed clay content is also recommended as a more appropriate boundary, informed by field-based and laboratory-based sediment dynamics, between ‘clean’ sand and ‘dirty’ sand, and between ‘arenite’ and ‘wacke’.