An experimental study on refugia condition for invertebrate (Isonychia japonica) by bluff bodies and its stability on the river bed

Abstract

Invertebrate behavior can be categorized into active walking, no walking and washout based on approach flow velocity. From this study, it can be concluded that invertebrates can actively walk for a depth averaged velocity less than 0.20m/s in a smooth bed condition (without roughness). Invertebrates showed no walking behavior at a depth averaged velocity more than 0.20m/s. Also, for a depth averaged velocity more than 0.40m/s, invertebrates were washed out. So, without roughness (WOR), the critical depth averaged velocity for no walking behavior is 0.20m/s and for washout behavior is 0.40m/s. It was also observed that invertebrates showed enduring behavior. The level of endurance was low during low flow conditions and high during high flow conditions. Artificial roughness was used at the upstream to increase turbulence in the flume channel and to understand the effect of turbulence on invertebrate. With the roughness (WR), critical depth averaged velocity for no walking is increased from 0.2m/s (without roughness (WOR) case) to 0.3m/s but critical depth averaged velocity for washout is decreased from 0.4m/s (without roughness (WOR) case) to 0.33m/s. The depth averaged velocity cannot estimate well the invertebrate behavior. So, local velocity at invertebrate height (4mm from bottom bed) was calculated by PIV. The results show that in a smooth bed (without roughness), local critical velocity is around 0.12m/s and 0.22m/s for no walking and washout respectively whereas in a rough bed (with roughness), local critical velocity is around 0.10m/s and 0.12m/s for no walking and washout respectively. The local critical velocity for no walking without roughness (0.12m/s) is similar to the value without roughness (0.1 m/s). For washout, the value without roughness (0.22 m/s) decreases when the roughness is introduced (0.12 m/s). So, the local velocity is found to be more important than depth averaged velocity for understanding invertebrate behaviors. Moreover, if depth averaged velocity is considered for invertebrate behavior analysis, then SCS ,which is spatially averaged shear component of turbulent intensity is responsible for invertebrate’s behavior of no walking and washout. However, if local velocity is considered, then with the introduction of roughness, at invertebrate height level, it can be concluded that local turbulent intensity, SCL which is related to shear or drag force that act along their body is responsible for invertebrate’s no walking behavior, whereas local turbulent intensity, VCL which is responsible for lifting invertebrate is more responsible washout of invertebrates. So, this study shows that the need for refugia is more under the turbulent condition of flow. Although this study doesn’t accurately predict the drift distance after dislodgement due to flume size limitation but the study shows that dislodgement of invertebrate is not increased with flow and the shear stress required to dislodge an invertebrate such as Isonychia japonica on a immoveable bed is higher without roughness (3.41 N/m2) than with roughness (0.12 N/m2). Different types of refugia were provided for invertebrates. Wooden blocks were used as refugia in the flume experiment. The two block setup with spacing (B) equal to height of the block (H) with small underscour and deep underscour at frontal block showed good results based on percentage invertebrate remain inside the flume. When refugia were provided, the invertebrates were able to walk for depth averaged velocity of 0.35 m/s that is 40% larger than that without refugia. With refugia, active walking and passive walking or no walking behavior was observed but washout behavior was not observed. With the introduction of refugia, the local velocity around the invertebrate height was reduced to less than 0.10m/s for high flow conditions. The upstream of block and gap inbetween blocks is more suitable in proving refugia than downstream of block for this type of setup. Water flow pattern between two neighboring blocks (refugia), the pressure distribution around the scoured block (refugia), the optimal spacing between two neighboring blocks (refugia) when underscouring of first block (refugia) occurs were analyzed. The flow pattern investigated between two neighboring blocks (refugia) with underscour shows that flow pattern was different for different horizontal spacings and underscour depths. It was noted that an eddy was not generated with small horizontal gaps between two neighboring blocks (refugia) and that larger eddies were generated with wider gaps. Further, when underscouring was deep, eddies were generated but were small in size compared to eddies generated when the underscour was shallow. The effect of wider horizontal gaps can be assumed to be significant in reducing the pressure on the bottom surface of a scouring block (refugia). For small and large underscours with a small horizontal space (B/H), the pressure distribution at the bottom surface decreases from front to back. However, for a deep underscour and a wide horizontal space (B/H=1), the pressure distribution was almost uniform at the bottom surface. Moreover, when the size of horizontal gaps was increased, the pressure acting on the top surface and front face was not altered much, while the pressure on the bottom surface was decreased and pressure on the rear face was increased. The drag and lift characteristics also explain the importance of wider spacing between two neighboring blocks during underscouring of the frontal block (refugia). Under the same underscour conditions, the lift and drag coefficient decreased when the horizontal gap between two blocks (refugia) increased. Moreover, with the same horizontal gap between two blocks (refugia), the lift coefficient increased and drag coefficient decreased with increasing underscouring depth. The results demonstrate that wider horizontal spacing between two blocks (refugia) is effective to prevent or reduce the possibility of collapse of the front block (refugia) when the underscouring becomes deeper. In addition, with shallow underscouring depths, blocks (refugia) seemed to achieve stability when gaps between two blocks (refugia) (B) were equal to height of the block (refugia) (H). Thus, the basic invertebrate behavior classification according to approach flow velocity is very important to understand the need for refugia during flood conditions and similar block arrangement (with horizontal spacing equal to height of the block) has a possibility to provide refugia in a stream. The flow pattern in-between two refugia and pressure distribution around scoured refugia has been clearly explained in this research. The blocks and its arrangement could be used as refugia during under scour conditions. The information in this research is useful for restoration projects in small stream for Isonychia japonica or similar Ephemerella species that have crawling behaviors and use small bed gravels to large rocks as refugia.