NonBasicSpecificComments05

• Title: Comments 5
• Date: 10/23/2003
• SWSI - Doc Type: Comments

10/23/200~ 10:40 FAX

I
~
t
t
1-
1

,
1
i
~
~
t
a
~
~
~
f

l.
!.
I
!
l

1
Ii
I
~
~

i
~.
4
)
~
5
s:
-:;
J
.;.
;;
i!
'1.
j
j
~
'~
~
~
~
{
~
5
i

. .

erali~ed model!? Lb~t are 'Widely available to predict
di schatge 1 n 'Watersh c-ds lacking gaging stations. Sur-
vjval estimates ~nd age and iTowth data are widely
avdilapl.t fOl' differ'ent trout species.
Our appl~oach acC"ourus for variation in stream hab-
itat over the lengrh of a stream or among streams In a
wat.ershed. The model of Hubert lilJ"ld others (1996)
acca un ted for variation in tr'ou t abundance mat was
primarily assoeiated with insn-cam cover. Use of aerial
photographs to estimate ~he amount of in~tream cover
in individual reach~5 ofJow"'gra.dicnr. streams appears' to
be a relatively accu:ra. te technique (Gtcentree and Al-
drich 1976. :Prlch~rd and o(hets 1. 99 6.. COvill.gton and
I-Iuberl 2000). Tlle condition of woody riparian vege-
tarion has heen shown to affect the a.bundante of in-
strC;Ltn cover for trout in several ttudles (Hubert 43.nd
others 1985~ PlattS an d others 1987 I \rVesche artd orb ers
· 1987, Marcus a.l1d. others 1990). SiJnUarly, the abun""
da.nce of ins-tream. cover has been relatt.::d to troUt bioI"
n'1as.s (Lewis 1969. DeVore and White 1978, Binn:; and
EiserOt3.n 1979, Wesche 1980)~ ~pA:cially brown trout
{Wesdle 1980, Sl1uler ~nd Nehring 1994~ Jowett Ig92s
1995). H'owever, it mlJst be ~tlecognized that restoranon
of insr.rea.m flows may al~o affect riparian vegetation
and inscream cover, thereby influeneiog the i.ndepen-
denol variables used to predict .a.bundarlce of ~f01)t*
The cumulative: effects ofvanous jnstream flows over
Inrge: portions of a watershed can b~ ~eBSed wing the
dtscrtbed approach~ For example, cffecr.s of va.rious
sum.mer flows on bro'Wn !fout biomass and abundance
in several trib\1[aries and the Upper Satt River were
e...;timatedi Because we approach is to be 1.1sed in an
iterative nlanner to identify the effi=C[S of different flows
on troUt biom:1SS and abundallce.. minimum it1st:reiUn
R01~ levels may be identified that produce an accept.,
able biomass or abundar.cc of trout while ma~ntaining
agricultul'"a] or municipal uses~ For eX:lTnple,. iterative
assessmenc.s for the Salt River Valley ldt:nufied the
stream segment with the greatest potential for produc-
ing ~he highest ilbundanc~ ofbruwn troUt 'Nith a givcn .
::tmounr ofWilter (i.c:., Upper Salt River).
Answers 1:0 questions. concerning trade-offs in agri.
c=ultural production :and fish production can be
wclghed usjng the approach. For example, We! com-
pa.....cd the use of water for production of alfalfa relative
to water oce~s to mai:nta.in bro'WO trOut in the sttldy
streams. .A.Jfalfa comprisc$ the m.PJor irrig3.red crop in
th~ valley (RavenhoJt and others 1976). The annual
r.unoun t of irrigation water net:d~d [0 raise alFalfa in the
Salt River Valley has been estima.ted to be 2715 m! jha
(Brosz 1997) ~ If 0.25 m~ Isec of W3.ter were converted
from .agricultu~ us3ge to instream flow, approxhnately
196 ha of la:od cO';lJd no longer be inigated. This is a

~ {Jl.J
p.l0

Trout Responses 10 Flaw RestOration 143

reiatively gm~n amonnl of ';Inn (0-42%) '-\lhc::,' COJH..
parelX to the 47.000 ha of irrigated land in th e &llt Rh'cr
Valll:Y (Sando .198.5). However. co retu."-n naLltJ~1 S\U1J"
mer flows to th~ Upper S~1t Rive.~ (1.0 m1- /~ec), 748 h~
of land could no longer be irrigated in the Salt River
Vall ey.
We ,=,stimat.ed brown trolH bio!n3SS: aJ1d 4lbUhd~YJce
in sU'cams.J but the estimates wert' limited by th e pTC~
djcrive models that we used. For example, the biomass
model Wa.~ highly significant (P < a,OOl) and h;.d been
tested an d validatr:d fol" broad application across Wyo-
ming (Hllbet~ and others 1996), bu[ the mode. ac-
cQunted for ::to limited amount of varia(ion in trOut
biomass (JtZ =: 0.40). Additiooall)'r the: model was not
built usi n g- da 1.il fro..n ~pring sa-earns or streams rhar
h3.~ populations of Ttu::'unr.ain whitefish,.. but flows in
one study srream were aftected by springs 'and moun-
rain whitefish occur in the Salt River and <:ould colonjzc
iSom e aib\ltatie9.o Consequen tly I the estim3.te$. of bio-
mass and abundanc:c nlust be assessed with knowledge.
[hat the prediccive modtls provide estima[e$; that hi:lve
substandal u:pcen.ainty assQctak'd ~th them.
Errors ca~l also occur when es.timating the indcpen--
den [ V3~riables used in the model LO predict U-O\11 bio...
m~s. Ell'eV~tion ;lnd channel ~di~nt estimates can be
afFc:cted by inaccurare mfcrp<:Jlations frQm tDpogr~phjc
maps. Estimates of wetted width were measured on site
during s'UfI1me.. base flow pCliod" and C~ be affected by
discharge at the: time of measurelneot. In~tream ct)ver
measuremel~lS are suJ:neetive (Scarnecnia and f;crg-
ersen 1987, Hogle and others 1995) and may bt:: J("ss
precise: wh en estimated from riparian fcanlres.
Accurate estimates of .A1lF under narural conditions
arc: an important c::omponen~ of our approach, Estimates \
ofADF can be made: in SC'VerAI different wa~ bu[ the b~t
data sources are strc3Jl~ gaug.es 'With long periods of
).ecotd. Howeve2:', gauged Streams a're not 3bun.d,:'iTH in
many 'Watersheds, For example, in [he Sall River V.alley
mere were fow- ~ugai strcoons with langwtcnn data, and
we extt'apoIir.ed etidmates to the other streams. Anor.he}~
approach to esr.i.rnadng ADF could be t.h e use of ma de]s
based on geomorph{c feawres of the warershcdtl For ex-
3nlpleJ Lowham (198B) :il1d Miselis .(1999) found th3~
bankfuU welted l'Jid r.b 'VfiU a signific.an t factor in p.redic:ting
discharge ttJ all regions of Wyoming.. However} when we
cOJ;npared cst;mares of ADF 1.15mg their modeb [0 recor.cls
from the gauged s~eams in the Salt rover drajnage~ their
model eStimates were considerably higher rh.:ln those
from gauge records.
We assumed a line: ar relati onsh ip b~tween trOUt nab--
itat and discharge for streams (Jowet.t 1997). but rhis
relationship rnay noC be at::r:uratc far all stream". Influ...
.:nee, of channel geomorphology can affe~t available