Linking local drift-feeding behavior to regional habitat envelopes using dynamic energetic gradients along a migratory corridor

Background/Question/Methods

Understanding how changes in local environments can impact regional populations of ESA listed species is a primary objective in conserving freshwater ecosystems.  In turn, regional mitigation efforts depend upon preliminary evaluations of habitat quality and restorative potential to rank alternative projects. Defining habitat quality is an important problem, yet preliminary restoration evaluations primarily rely upon subjective professional consensus and/or a disproportionate reliance upon indirect environmental gradients, the latter of which may have no direct physiological relevance to the focal species. Here we apply a quantitative means to bolster habitat evaluation efforts by integrating an energetically driven movement model with a dynamic hydraulic model. We begin by predicting per capita intake rates experienced by a drift-feeder and extrapolate to regional distribution densities and residency times across habitats in the Lower Columbia River (LCR).  Model predictions are compared to both laboratory and field data.

Results/Conclusions

Results show that gradient-oriented foraging effectively predicts observed mean per capita capture rates under a broad array of flow conditions, while over-estimating capture success at high flows.  However, optimal intake rates between select species occur below aerobic swimming thresholds (e.g., critical swimming speeds), which fall within moderate flow regimes. The movement model is then coupled with regional flow patterns generated for the LCR between 2007-2009 and the resulting velocity fields are converted into dynamic energetic gradients based on species-specific swimming costs. Individuals follow optimal flow regions to minimize travel costs across two freshwater habitats and the residency time and local densities are compared to field data to rank alternative restoration efforts.

In conclusion, Brownian motion coupled with an optimal energetic cost function can effectively predict per capita gains under a wide range of bounded flow conditions.  Critical swimming speeds provide an effective threshold to bound our predictive range and effectively reduce the uncertainty associated with predicting growth potential in drift feeding salmonids.  Replication of local, fine grain behavioral trends can then be scaled up and applied to regional habitats along an endangered species’ migratory corridor. The result is a quantitative means of reducing preliminary uncertainties in ranking alternative restoration efforts.