|
<br />o v"th~ r !'v~O..l^t} heu,
<br />~,c -, ,Tq'b~ J
<br />
<br />..'
<br />
<br />(3) Always check the fit of the current model through resid-
<br />uals and other diagnostics. Again. Fournier's point about
<br />checkiny NsiL/uuls III w!:1! rllk!:n - S. J. Smith. Mew/l", Fi.'"
<br />Dil'isiol/. 8edJilt'd II/stitllte of O(,(,(l1/ography. P.O. 80x /006,
<br />Dartmourh. N.S. 82Y 41\2. and Phillip E. J. Green, COl/sllltil/g
<br />Statisticial/. Departmem of Mathematics. Staristics and Com-
<br />putil/g Science. Dalhollsie Uninn;,.y. Hal!fax. N.S. 83H 4H4.
<br />(J8515)
<br />
<br />A cknOll'1 edgeml'llt.~
<br />
<br />We would like to thank Dr. Doug Wiens and Mr. L. P. Fanning for
<br />reviewing the manuscript.
<br />
<br />References
<br />
<br />HILBORN R., AND M. LJ::DBEITER. 1985. Delerminanls of catching power in
<br />Ihe British Columbia salmon purse seine nee!. Can. J. Fish. Aqua!. Sci.
<br />42: 51-56.
<br />McCULLAGH. P., AND J. A. NELDER. t983. Generalized linear models.
<br />Chapman and Hall. London. 261 p.
<br />SAS INSTI11JTE INC. 1982. SAS users guide: slalistics. 1982 ed. SAS Institute
<br />Inc.. Cary. NC. 584 p.
<br />SEARLE, S. R. 1971. Linear models. John Wiley & Sons, Inc., New York. NY.
<br />532 p.
<br />WHITMORE, G. A., AND J. F. GENnEMAN. 1985. Iceberg paths and collision
<br />risks for fixed marine structures. Can. J. Sta!. 13(2): 83-108.
<br />
<br />.J,p-V ~;
<br />
<br />U1 In Defense of the Instream Flow Incremental
<br />Methodology
<br />
<br />Malhur et al. (1985) recently criticized the Instream Flow
<br />Incremental Methodology (IFIMl. a technique for recommen-
<br />ding flows for streams, on the basis of the evidence available
<br />to support the main assumptions. The major point of their
<br />article was that the critical assumption of a positive relationship
<br />between weighted usable area (WUAl. a measure of usable
<br />habitat, and biomass of fish ....as not well established. While
<br />this concern and others expressed'by Mathur et al. (1985) are
<br />legitimate and should be addressed in continuing research
<br />and methodology development, the article contained several
<br />misinterpretations of previously published results. Our com-
<br />ments deal with three topics: (I) suitability index curves,
<br />(2) habitat-biomass relationships, and (3) validation studies.
<br />The suitability curves depicted in fig. I of Mathur et al.
<br />(1985) were used to demonstrate that curves developed from
<br />different streams can vary in shape and location. However, the
<br />curves shown were developed using only habitat utilization
<br />data and adjustments were not made for habitat availability at
<br />time of sampling. Baldridge and Amos (1982) and Orth et al.
<br />(1982) demonstrated the importance of ad'ustio 0 a it at
<br />availability at time of sam lin. a It'y,,=_n'es a e ,~
<br />n I Izal 1l..; !s. !lit
<br />A pOSitive relationship between WUA and fish biomass or
<br />numbers is assumed in any IFlM application. The evidence to
<br />support this relationship indicates that the relationship holds
<br />only during periods of limiting habitat. Furthermore, time lags
<br />between habitat limitation and population response complicate
<br />interpretation of the nature of the relationship in unregulated
<br />streams. Mathur et al. (1985) plotted our data (Orth and Mau-
<br />ghan 1982) on seasonal biomass and WUA in their fig. 3 and
<br />4 to support their conclusion that no positive, linear re-
<br />lationship exists. Their plots are misleading, since data from all
<br />seasons were pooled: each plotted datum represents a weighted
<br />average from four sites (two pools and two riffles). Further-
<br />more, one datum in fig. 4 (freckled madtom: 199 m2, 2 g) was
<br />
<br />1092
<br />
<br />0'; Cf'3L
<br />
<br />-
<br />
<br />not even in our data set. It was unreasonable of these authors
<br />to expect any relationship when data were plotted in this man.
<br />ner, Our hypotha!\ls hUll been thot ulloblo hubltut may IIrnlt
<br />population biomass only during limited time periods. While
<br />this can occur in any season, we found that biomass of three
<br />riffle-dwelling fish species was correlated with thei'r respective
<br />measures of WUA for the summer only (Orth and Maughan
<br />1982). In winter and spring, WUA was usually higher but fish
<br />populations did not respond because recruitment of young fish
<br />had already been limited by summer conditions. This hypoth-
<br />esis was later supported by Loar et al. (1985) who found
<br />that brown trout (Salmo truua) biomass was correlated with
<br />the minimum WUA for adults over the year (R2 = 0.71;
<br />P = 0.0001) and rainbow trout (Salmo gairdneri) numbers in
<br />allopatric populations were correlated with the minimum WUA
<br />for incubation (R2 = 0.62; P = 0.000 I). These results em-
<br />phasize that available habitat can limit fish populations during
<br />relatively short time periods. However, available habitat is not
<br />always positively related to population abundance due to other
<br />potential limiting influences, such as interspecific competition
<br />(Loar et al. 1985). These complicating factors do not detract
<br />from the usefulness of IFIM but they do point to the need for
<br />more research on the influence of stream flow on stream fish
<br />p$Ulations and communities.
<br />, ~"'m ..~ f1F
<br />~ .9.ngorous y" s ~'...__ ' e '.
<br />detennfne--stffi<Oo\\l"" s eets CI. .~.
<br />Imp.e~~...~.',~QP1men. e...... .0. W'
<br />"lockS 0 IS '(' -...~<" "'.~ _ . pon~'
<br />t-1.m~
<br />.jJIJ -
<br />
<br />a.~..",......~ll~~_!!lJ~r;. s
<br />(1985) clteCl the study by Kraft 97 as eVI ence that reduc-
<br />tion in summer flows altered only distribution and not abun-
<br />dance of brook trout (Salvelinus fominalis). However. in
<br />Kraft's (1972) study, changes in physical characteristics of the
<br />study sections after a 90% reduction in flow were not neces-
<br />sarily detrimental because the study stream had a well-defined
<br />channel and the reductions in available habitat were not in
<br />proportion to flow reductions. Since availability of usable area
<br />was not quantified, it was wrong to conclude, as Mathur et al.
<br />(1985) did, that usable area may not have a regulatory effect on
<br />the population. In conclusion, while we agree that further study
<br />is needed, we support the current practice of making the
<br />assumption that WUA is an index of potential Fish biomass or
<br />numbers - Donald J. Orth, Department of Fisheries and Wild-
<br />lifeSdences, Virginia Polytechnic Institute and State Univer-
<br />sity, 8lacksburg, VA 24061, USA, and O. Eugene Maughan.
<br />Oklahoma Cooperative Fish and Wildlife Research Unit,
<br />Oklahoma State University, Stillwater, OK 74078, USA.
<br />(J844la)
<br />
<br />
<br />References
<br />
<br />BALDRIDGE, J. E., AND D. AMOS. 1982. A lechnique for determinil1g fish
<br />habitat suitability crileria: a comparison between habitat utilization and
<br />availability, p. 251-258. In N. B. Armantrout [cd.] Symp.. Acquisition
<br />and ulilization of aqualic habital inventory information. American Fish-
<br />eries Society, Bethesda. MD.
<br />KRAFT, M. E. 1972. Erfecls of Cl.'ntrolled now reduction on a Iroul stream. J.
<br />Fish. Res. Board Can. 29: 1405-14tl. .
<br />LoAR. J. M., M. J. SALE, G. F. CADA, D. K. Cox, R. M. CUSHMAN. G. K.
<br />EDDLEMON. J. L. ELMORf-. A. J. GATZ. P. KANclRUK. J. A. SOLOMON.
<br />AND 0, S. VAUGHAN. 1985. Applicalion of habilat evaluation models i.n
<br />southern Appalachian troul streams. Oak Ridge Natl. Lab. Environ. S':I.
<br />Div. Publ. No. 2383, ORNL./TM-9323. Oak Ridge. TN. 310 p.
<br />
<br />Can. ~..F;.fh. Aqual. Sci.. Vol.4J.J9H6
<br />
|