Limits on salmon spawning in coarse-bedded rivers

Background/Question/Methods

Revitalizing salmon spawning in anthropogenically coarsened riverbeds is a top objective and expense of river restoration.  However, the influence of coarse sediment on salmon spawning has not been systematically quantified.  Moreover, there is no readily applicable tool for determining when riverbeds become too coarse for spawning.  Hence, managers have had little quantitative basis for selecting sites that would optimize benefits of gravel augmentation for salmon spawning, making such restoration projects prone to inefficiency and ineffectiveness.  To quantify the physical limits on salmon spawning in coarsened riverbeds, we studied grain-size distributions in redds and the areal extent of spawning in a series of California and Washington rivers where salmon spawning ability appears to be affected by coarse bed material. 

Results/Conclusions

We show that spawning decreases with increasing coverage by particles that are too coarse to move during redd building by chinook (Oncorhynchus tshawytscha) and pink salmon (O. gorbuscha).  Interspecific differences in average mass-to-length ratio (MLR) explain differences in mass of the largest movable particle in redds.  Moreover, we observe interspecific differences in the relationship between spawning use and areal coverage of immovable particles; for pink salmon, spawning use decreases from 92% to 5% as percent coverage of immovable particles increases from 20 to 60%, whereas, for chinook salmon, spawning use decreases from 35% to 2% as percent coverage of immovable particles increases from 15 to 50%.  We present a simple, quasi-mechanistic regression model that predicts percent coverage by immovable particles to within 10%, based on MLR and readily measured particle-size indices (i.e., D₈₄ and the geometric standard deviation).  Dependence on MLR suggests the model captures interspecific differences in ability to move sediment.  When coupled with our data on spawning use, our regression model enables us to predict the suitability of coarse gravel for different fish species.  This approach represents a generic, readily applicable tool for use in quantifying, managing, and restoring salmonid spawning habitat in rivers with coarse beds.