Bormann 1974). However, for many years, the importance of out-of-bank flows to the riverine

ecosystems was not recognized by instream methods used to assess flow requirements

(Middleton 1999). Methods such as the Tennant method or the use of wetted perimeter can be

serviceable in the protection of the lowest flows. However, they deal predominately with

between bank analyses and offer little to no protection for high flows, though both methods have

been widely applied (Gippel and Stewardson 1998, Jowett 1997, Postal and Ricther 2003).

While these methods offer valuable information for regulating the low end of the flow regime, it

is necessary to measure and assess the flows required at the high end of the flow regime as well.

 Munson and Delfino (2007) state that quantification of potential differences in floodplain

inundation associated with differences between minimum flow requirements and natural

hydrologic regimes should be a critical component of regulatory programs that are intended to

protect or restore riverine habitat. The District has employed a method for establishing MFLs

that acknowledges the importance of seasonal flow regimes, including flows necessary for

floodplain inundation. The method employs the historic flow records and a temporal measure of

habitat loss based on the reduction in the number of days specified flows are reached under

various flow reduction scenarios. An alternative approach, employing a measure of spatial loss

that examined the change in top width at each cross section modeled under different flow

conditions, was previously explored in this dissertation (Munson and Delfino 2007).

 Both spatial and temporal measures of habitat loss can be related to percent flow reduction

in a river channel. However, as previously discussed, both serve as proxies to evaluate the extent

of the river-floodplain connection. This connection has not yet been fully described on any river

studied by the District because complicated floodplain geometry is being represented by a

limited number of historic floodplain cross-sections derived from earlier studies. Though these