Hovering in turbulent rock-array microhabitats: Multidimensional kinematic responses of Schizothorax grahami

The interaction between flow and riverbed structures creates spatially heterogeneous turbulent microhabitats that influence fish station-holding strategies and energy allocation. This study investigated the hovering behaviour of Schizothorax grahami in complex turbulent microhabitats formed by rock arrays of different scales using flume experiments. Video analysis and data storage tag (DST) tilt sensors were combined to quantify five multidimensional kinematic indicators: ground swimming speed ( $V_g$ ), angular velocity of swimming direction ( ${\omega }_{\alpha }$ ), tail-beat angular velocity ( ${\omega }_{\beta }$ ), pitch angular velocity ( ${\omega }_{\theta }$ ) and roll angular velocity ( ${\omega }_{\phi }$ ). As rock-element size increased, fish hovering positions shifted from dispersed distributions to regions with local velocity gradients around rock elements, with clear positional overlap observed under some large-rock-array conditions. Increasing discharge generally increased the mean values and standardized within-condition variances of most kinematic indicators. Trial-level Spearman correlation analysis showed that $V_g$ had the strongest positive correlation with ${\omega }_{\alpha }$ . Generalized linear mixed models further revealed functional differentiation among kinematic responses: $V_g$ and ${\omega }_{\alpha }$ were mainly driven by flow velocity ( $U_{\mathrm{mag}}$ ) and turbulent kinetic energy, whereas ${\omega }_{\beta }$ , ${\omega }_{\theta }$ and ${\omega }_{\phi }$ showed weaker hydrodynamic responses. These findings suggest that fish hovering is an active multidimensional regulation process in which position maintenance, directional correction, propulsion and postural stabilization show partially differentiated response patterns. These findings provide new insights into the locomotor and postural strategies that enable benthic fish to maintain stability in heterogeneous turbulent environments and improve our understanding of fish behavioural adaptation to complex flows.