Crab burrow aspect ratio influences particle capture rates on intertidal sandflats

Abstract

Crabs are ecosystem engineers that dig burrows, which can trap sediment. We used a field experiment to explore how burrow aspect ratio (depth/diameter) affects the trapping efficiency of sediments and organic material. Arrays of burrows mimics were constructed using tubes of similar depths but two different opening diameters: fat tubes (aspect ratio of 3.8) and thin tubes (7.1). Different arrays were tested to examine whether the combination of burrows with two different aspect ratios affects the material capture, while retaining the same total area of openings per array. The results showed that, in general, the fat tubes trapped more organic material, especially large pieces, and more sediment than thin tubes. Furthermore, the silt-clay content of the sediment trapped in the tubes was up to 50% greater than the surrounding surficial sediment. Hence, we conducted 2D numerical simulations of flow around, and into, a single burrow to elucidate the mechanisms behind particle capture. Results showed that the flushing rate and the turbulent kinetic energy were greater in the fat tubes. However, turbulence persisted for a longer distance downstream of the thin tubes than the fat tubes. The enhanced turbulence may increase the flux of sediment into the tubes where particles can settle and, consequently, promote the sediment capture and sedimentation rate. Our results demonstrate that the aspect ratio of burrows significantly affects capture processes of organic material and sediment particles. Moreover, this capture mechanism depends on the flow regime, as faster flows are associated with larger amounts of material in the water column.

Model setup and validation

In order to verify and validate the numerical model, several benchmark tests were conducted. According to Oberkampf and Trucano (2008), the simulation of a same case should be run for three different grid sizes as well as three different time steps. In addition, numerical data should be compared with experimental data (numerical, flume or field) to ensure that the model output corresponds to the reality and that its accuracy is in an acceptable or reasonable range. A large number of studies have already been published regarding the flow in cavities: 2D or 3D lid-driven cavity flow in square/cubic or rectangular cavities (Ghia et al. 1982; Ku et al. 1987), flow past open cavities with different shapes (Ozalp et al. 2010). In order to validate our simple model, a square unit cavity and a rectangular cavity were generated. The geometry, as well as the boundary conditions, is shown (Fig. 9a). In total, a set of nineteen numerical experiments was conducted. Nine Reynolds numbers were considered, ranging from 100 to 100,000. A vertical profile of mean streamwise velocity was extracted at X/D = 0.5 and an horizontal profile of mean vertical velocity was extracted at Z/H = 0.5, where X and Z refer to the horizontal and vertical axes, respectively, and, D and H are the length and the height of the cavity, respectively. These profiles were plotted against data from Ghia et al. (1982), Takemoto et al. (1984), Babu and Korpela (1994), Cortes and Miller (1994) and Botella and Peyret (1998). The results of our model validation exhibited good agreement with these numerical data for mean flow speeds and vortex structure (Fig. 9b, c).

Fig. 9

Geometry and flow validation of hydrodynamics in square cavity at Re = 1000. a Boundary conditions and geometry of the square cavity (height H equal to diameter D). b Modelled results of mean streamwise velocity plotted against published data. Data were extracted along a vertical profile located at X/D = 0.5. c Modelled results of mean vertical velocity plotted against published data. Data were extracted along a horizontal profile located at Z/H = 0.5