7926

• UCREFRP Catalog Number: 7926
• Author: Annear, T. C. and A. L. Conder.
• Title: Relative Bias of Several Fisheries Instream Flow Methods.
• USFW Year: n.d.
• USFW - Doc Type: Cheyenne, Wyoming.
• Copyright Material: NO

tively) provi.ded significant numbers of unbiased estimates.

, ,'~~;J!;f ~
::':~,i~:;,
Of the recollUllen...
u~-~ .'

dations that were biased for these two methods, all three were biased upwards

for Method 10 and all three for Method 14 were biased low. This information

sugges t that Method 10 1S the preferred of the two as, if the estimates

later found to be in error, overestimates are often more acceptable

fisheries than underestimates, unless they are unreasonably high.

Neither of the two methods incorporating recommendations for all habitats

(~ethods 9 and 12) was significantly unbiased, however, Method 12 (where two of

three criteria were met) was biased low f.or 12 of 13 streams.

This tendency

creates a high level of risk that an instream flow recommendation could under-

estimate fishery needs. Method 9, generated eight unbiased recommendations for

the 13 streams with the five biased recommendations being almost equally divided

between overestimates and underestimates of unbiased ranges. Though not signi-

ficantly unbiased, this method would probably be of greater utility than Method

12 where users wish to address instream flow needs in all habitats within a

stream segment.

An associated drawback with this general category of tnethods is the dif-

ficul ty in identifying ~'cri tical riffles" in the field, especially when the

first field visit occurs during high water.
I

This is also a problem where

distinct differences between riffles are not apparent at low flows. To minimize

this problem, investigators should collect data for at least three riffles in

each stream reach and select the critical riffle after completion of data

analysis.

This approach offers several advantages for obtaining instream flow recom-

11