Movement is a fundamental population process, but a relatively new topic in ecology and poorly understood. While studies recognize the importance of population density and social forces on movement within and between habitats, management of aquatic habitat is often limited to alternatives in geometry, flow, and water quality. Few tools directly relate these managed attributes to movement in a manner recognizing the importance of limited information at the individual level due to varying sensory acuity and acclimatization based on past experience, changes in internal state, irregularity and partial preferences in decisions, interactions between intrinsic and extrinsic (stimulus-related) factors affecting motivation, and continuous behavior adjustments in response to these factors. Fish are not primitive in these respects and ecological models are inadequate without reference to these cognitive processes. Using field movement data we demonstrate the utility of a Eulerian-Lagrangian-agent method (ELAM) for analysis of the process-based relationships between patterns in movement and managed environmental stimuli. Simply, an ELAM is an individual-based model (IBM) integrating a particle-tracking model (PTM), well-established models of the physics (hydraulics) and water quality of the aquatic environment, and emerging theory in cognitive ecology describing how animals perceive and respond to multiple environmental stimuli through time.
Results/Conclusions
ELAMs have been applied and validated at detailed and field scales. On the Columbia and Snake rivers, ELAM analysis yielded a process-based description for how juvenile salmon use multiple hydraulic gradients to “image” the hydrogeomorphology of natural rivers for navigation, without need for vision. When these navigation rules for natural rivers are “played” within the modified flow regime of a hydropower dam it is possible to describe the detailed (sub-meter, 3-D) approach paths of acoustically-tagged salmon as well as their passage through five hydropower dams (spillway, r2 = 0.87; fish bypass, r2 = 0.81; powerhouse turbines, r2 = 0.49), consistent with the flow variability of these routes. In 60-km long J. Strom Thurmond Lake (Georgia–South Carolina), ELAM analysis yielded a process-based description of the temperature-dissolved oxygen “squeeze” experienced by landlocked blueback herring (Alosa aestivalis) describing longitudinal (r2 = 0.67) and vertical (r2 = 0.93) distributions of fish observed through mobile hydroacoustics and species composition data from gillnets. The longitudinal distribution of fish was explained by a random term (57%) and horizontal water velocity (43%) while the vertical distribution was explained by water temperature (45%), a random term (44%), dissolved oxygen (6%), horizontal velocity (4%), and vertical velocity (1%).
