Surrogates of juvenile kelp were used to assess how they affect the critical shear velocity of the stones they are attached to. For this purpose laboratory experiments were carried out in April 2023 at the Ludwig-Franzius-Institute of Hydraulic, Estuarine and Coastal Engineering. Stones were exposed to increasing unidirectional flow and the experiment was terminated when the stones started to move. Kelp surrogates comprised of an artificial kelp attached to individual stones. For the stones all three dimensions were measured with a calliper gauge and their weight was determined with a balance to the nearest 0.01 g. Their volume was determined by water displacement. The artificial kelp was produced from a geotextile (thickness = 0.8 x 10⁻³ m, density = 258.1 kg/m³, flexural rigidity = 110.55±45.26 x 10⁻⁴ Nm²). They were cut in rectangles of 2 x 10 cm and one short side was glued to the stones with superglue. The kelp surrogates were individually exposed to unidirectional flow in a circulating flume with a straight working section of 1.13 m length, 7.7 cm width and 15 cm height. Flow was controlled via an inlet valve and a sharp crested weir at the end of the working section. Surrogates were placed in the centre of the working section and an Acoustic Doppler Profiler (ADV, Vectrino Profiler by Nortek) was used to measure flow velocities 45 mm in front of the surrogates. It was set to measure flow velocities at 100 Hz in 1 mm bins within the first 10 mm above the bottom. The flow was increased incrementally by adjusting the weir height and discharge; each flow velocity was maintained for 3 minutes during which the motion of the kelp surrogate was observed. If it remained stable during this time, the velocity was increased. If it moved, the previous velocity was defined as critical. Consecutively, the length of the artificial kelp was reduced by 2 cm and the experiment started again with a low flow velocity. This process was repeated until only the stone remained and its critical flow velocity was determined. From the ADV data, only the downstream velocity component was considered. Poor quality data was removed from the time series by excluding data points with a Signal-to-Noise-Ratio <15 dB and a correlation <90 %. The remaining data was used to compute a mean value for each position in the velocity profile. Visual inspection showed that vertical velocity distributions followed a logarithmic profile in all cases. Subsequently, data from 1 cm above the bed was used to compute the critical shear velocity u_crit according to: u(y)/u_crit = 1/K * ln(u_crit * y / v) + 5.5 With u(y) being the mean velocity at height y, K = 0.4 being the von Kármán constant and v being the kinematic viscosity (1.002 * 10⁻⁶ m²/s).