Beamer, E.M. and Henderson, R.A., 1998. Juvenile Salmonid Use of Natural and Hydromodified Stream Bank Habitat in the Mainstem Skagit River, Northwest Washington. Skagit System Cooperative, La Conner, WA. pp. 55.

This study examined juvenile salmonid use and habitat changes associated with
stream bank protection at the site level in the mainstem Skagit River. Natural and
hydromodified banks were paired by location over an eighty mile-river length. Bank habitat
was defined as either “natural” or “hydromodified” based on the presence of riprap or other
human induced bank modification and distinguished from mid-channel units ( e.g., riffle,
glide, and pool) by differences in water current velocity.
Natural banks had ahigher percentage of their area in wood, cobble, boulder,
aquatic plants, undercut bank, and no cover types when compared to hydromodified banks.
, • While no riprap or rubble was foun&in natural banks, wood .cover~was,common.in some
hydromodified banks. Wood cover in hydromodified banks increases with increasing time
after hydromodification. We found no significant difference in water surface velocity
between natural and hydromodified banks. Riprap/rubble and wood cover were not
correlated with water surface velocity. However, the gradient of the bank unit and the
streamflow discharge were correlated with water surface velocity.
The findings of this study reveal some consistent trends in fish use across sampled
reaches. For juvenile chinook and coho in bank habitat, fish abundance has a significant
positive correlation with the amount of wood cover. Wood cover jn hydromodified banks
explained 82% of the variation in chinook abundance. For juvenile coho at the end of
summer rearing, wood cover in both bank types explained 62% of the variation in fish
abundance. There is evidence of preference for riprap (but not rubble) and some specific
types of wood cover by rainbow suggesting that rainbow may not be adversely impacted
at the site level by bank hydromodification if rock particles are large. While wood cover
is the most common natural bank cover, fish abundance within wood cover types is not
uniform. Fish abundance is greater in rootwad cover than single logs for all species and
life stages examined, except sub-yearling chum. Sub-yearling chum prefer aquatic plants
and cobble, two other cover types more common in natural banks. The findings also
suggest that the use of natural cover types along with bank protection may mitigate some
site (but not reach) level impacts ofhydromodification. Results presented can form the
basis for estimating restoration project benefits, planning mitigation, or assessing the
impacts of habitat loss.