Biological Services Programand

Division of Ecological ServicesFWS/OBS-82/10.25SEPTEMBER 1982

HABITAT SUITABILITY INDEX MODELS:REGRESSION MODELS BASED ONHARVEST OF COOL AND COLDWATERFISHES IN RESERVOIRS

SK lh and Wildlife Service361

~~~482- S. Department of the Interior10.25

The Bi010g1 cal Services Program was es tab l1shed wi thin the U.S. Fi shand Wildlife Service to. supply scientific information and methodologies onkey environmental issues that imp~ct fish and wildlife resources and theirsupporting ecosystems. The mission of the pr.ogram is as follGWs:

• To strengthen the Fish and Wildlife Service in its role asII primary source of information on 0lt10nal fish and wild­life resources. particularly in respect to environmentalimpact assessment.

• To 9lther. Inalyze. a·nd prese"t tnformation that will aiddecisiOJllllaktrs in the id·entification and resolution ofproblems associated with maJ;orchanges in land and wateruse.

• To pro.vide better ecol09iC:a1 inforlllltionand evaluationfor Department. of the Interior development programs. suchas those relating to energy development.

Information developed by the Biological Services Program is intendedfor use In the planning and d.tisionlllJking proc:.ss to prevent cr minimizethe impact of development O~ fish .nd wildlife. Research activities andtechnical assistance servi~s ere blsed on an analys1S of the issues, adetermination of the decis;onmakerS iny~lved and thei~ information needs.and an evaluation of the state of the art to identify information qlpsand to determine priorities. T~is is i strategy that will ensure thatthe products p~oduced and dtisem1na&ed ar& t~ly and useful,

Projects hay! been i.flit1.~d HI the fQllowin.g areas: coal extractionand conversion; power plants,geotht.f'flII1. 1Ill1.f}~ral and 011 shale deve lop­ment; water resource ~nal1Si$~ fnc'l\lGififf .!icts:'llilIillterations and westernwater alloeation; C'Guta.l ee-Q$,y$tllll$i1:M Outer GonUnental Shelf develop­ment; and sys t8111S inventory, 1t1!1lldtDgftatlona1 Wet1and Iflvento ry ,nab; tat class ; f~ cat'ion «tid' analys;h, al'ld 1tflfbrlllation traMfer.

Th,. BicilOgh:.l $e'",itP PtIOgTPl t_f.~.¢ t"'_ 'O·fflee of BiologicalSer\liee:!i tnWl:sh1l\gttoD~ D..~ .• ",*M~ 1s.rnp.p,rsnti'le fol' ttv'rall planning andmanageJJte1tt; NiUona1'1'.." 'wll1_lJr~1'4' tJte Program's c,entral scientificand t.~"nlc.' ••pe~tI ~ .r¥4~ ~r ecntt~ting biolo~ical servicesstudies "ittl Hates~ Uf:l~S'ftl",\ COnif.l1\fngf'frms, and others; RegionalStaffs. wtw provide a litl&: 'W IlMlems at 'be op.e:ra;ting level; and staffs atcertain Fish and Wildlife Service ....uln:h ftE:'flHies • .,ho conduct in-houseresearch stu.dil'S.

This model ;s designed to be used by the Division of Ecological Services1n conjunction with the Habitat Evaluation Procedures.

FWS/OBS-82/10.25September 1982

HABITAT SUITABILITY INDEX MODELS: REGRESSION MODELS BASED ONHARVEST OF COOLWATER AND COLDWATER FISHES IN RESERVOIRS

by

Larry R. Aggus and William M. BivinU.S. Fish and Wildlife Service

National Reservoir Research ProgramFayetteville, AR

Project Officer

James W. TerrellWestern Energy and Land Use Team

2625 Redwing RoadFort Collins, CO 80526

Performed forWestern Energy and Land Use Team

Office of Biological ServicesFish and Wildlife Service

U.S. Department of the InteriorWashington, DC 20240

This report should be cited as:

Aggus, L. R., and W. M. Bivin. 1982. Habitat suitability index models:Regression models based on harvest of coolwater and coldwater fishes inreservoirs. U.S. Dept. Int. Fish Wildl. Servo FWS/OBS-82/10.25. 38 pp.

PREFACE

The methods presented in this report are designed to permit habitatclassification of reservoirs, containing coolwater, coldwater, and seasonaltwo-story fisheries, based on harvest of selected coolwater and coldwatersport fishes. Multiple regression equations describing relations betweenreservoir environmental characteristics and biomass harvest of selected sportfish species or groups are presented. Cumulative Frequency (CF) plots ofknown harvest estimates from the various classes of reservoirs are presentedto facilitate conversion of harvest predictions to Habitat Suitability Indices(HSI's). Detailed descriptions and limitations of the procedures are dis­cussed.

The predictive capability of the regression equations is likely to varydepending on the environmental conditions present. The U.S. Fish and WildlifeService encourages model users to convey comments and suggestions that mighthelp us increase the utility and effectiveness of this habitat-based approachto planning. Please send comments to:

National Reservoir Research ProgramU.S. Fish and Wildlife Service100 West Rock StreetFayetteville, AR 72701

or

Habitat Evaluation Procedures GroupWestern Energy and Land Use TeamU.S. Fish and Wildlife Service2625 Redwing RoadFt. Collins, CO 80526

iii

CONTENTS

PREFACE iiiFIGURES vTABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. vi iACKNOWLEDGMENTS vi i i

INTRODUCTION 1HABITAT SUITABILITY INDEX MODELS.... 2

Model Applicability.......... 2Selection of Variables................................................. 2Data Analysi s 3Interpret i ng Model Outputs 11

RECOMMENDATIONS 12

REFERENCES 13APPENDICESA. Definitions of Environmental Variables Used in Equations for

Predicting Harvest of Coolwater and Coldwater Fishes inReservoi rs 15

B. List of Reservoirs and Years of Harvest Record Used to DevelopPredictors of Harvest in Reservoirs with Coolwater, Coldwater,and Seasonal or Two-story Coldwater Fisheries .. ~. 16

C. Multiple Regression Formulas for Estimating Harvest (kg/ha)of Coolwater and Coldwater Fishes from Reservoirs......... 20

D. Cumulative Frequency Plots of Harvest of Selected Sport Fishesin Reservoirs with Coolwater, Coldwater, and Seasonal or Two-StoryColdwater Fi sheries 23

iv

Number

1

0-1

0-2

0-3

0-4

0-5

0-6

0-7

0-8

0-9

0-10

0-11

0-12

0-13

FIGURES

Cumulative frequency plot of walleye harvest. showing aprocedure for obtaining CF values .

Cumulative frequency plot of walleye harvest inreservoirs with coolwater fisheries .

Cumulative frequency plot of sauger harvest in reservoirswith coolwater fisheries .

Cumulative frequency plot of yellow perch harvest inreservoirs with coolwater fisheries .

Cumulative frequency plot of northern pike harvest inreservoirs with coolwater fisheries .

Cumulative frequency plot of muskellunge harvest inreservoirs with coolwater fisheries .

Cumulative frequency plot of Esox spp. harvest inreservoirs with coolwater fisheries '" .

Cumulative frequency plot of rainbow trout harvest inreservoirs with coldwater fisheries .

Cumulative frequency plot of brown trout harvest in reservoirswith coldwater fisheries .

Cumulative frequency plot of cutthroat trout harvest inreservoirs with coldwater fisheries .

Cumulative frequency plot of brook trout harvest inreservoirs with coldwater fisheries .

Cumulative frequency plot of total trout harvest inreservoirs with coldwater fisheries .

Cumulative frequency plot of lake trout harvest inreservoirs with coldwater fisheries '" .

Cumulative frequency plot of coho salmon harvest inreservoirs with coldwater fisheries .

v

10

23

24

25

26

27

28

29

30

31

32

33

34

35

FIGURES. (concluded).

Number Page

0-14

0-15

0-16

Cumulative frequency plot of kokanee salmon harvest inreservoirs with coldwater fisheries .

Cumulative frequency plot of total trout harvest inreservoirs with seasonal or two-story fisheries .

Cumulative frequency plot of rainbow trout harvest inreservoirs with seasonal or two-story fisheries .

vi

36

37

38

Number

1

2

3

4

5

B-1

B-2

B-3

TABLES

Fish harvest (dependent) variables and environmental(independent) variables used to develop habitatsuitability indices for coolwater and coldwaterfi shes in reservoi rs .

Number of observations, means, ranges, and standarddeviations for selected environmental and sport fishharvest variables from reservoirs with coolwater fisheries .

Number of observations, means, ranges, and standarddeviations for selected environmental and sport fishharvest variables from reservoirs with predominantlycoldwater fisheries .

Number of observations, means, ranges, and standarddeviations for selected environmental and sport fishharvest variables from reservoirs with seasonal ortwo-story coldwater f i sheri es .

Classification of walleye and rainbow trout reservoirs,based on observed groupings of estimated annual harvest .

Reservoir names and numbers of years of harvest recordsused to develop suitability indices for coolwater fishesin reservoirs .

Reservoir names and numbers of years of harvest recordsused to develop suitability indices for coldwater fishesin reservoirs with predominantly coldwater fisheries .

Reservoir names and numbers of years of harvest recordsused to develop suitability indices for coldwater fishesin reservoirs with seasonal or two-story coldwaterfi sheri es .

vii

4

6

7

8

11

16

18

19

ACKNOWLEDGMENTS

We would like to thank. Bill McConnel, Sam Williamson, and Dick Applegatefor their review comments. Word processing was provided by Carolyn Gulzow andDora Ibarra. Editing was provided by Cathy Short. The cover of this documentwas illustrated in part by Jennifer Shoemaker, and in part with material takenfrom Freshwater Fishes of Canada, Bulletin 184, Fisheries Research Board ofCanada, by W. B. Scott and E. J. Crossman.

vii i

REGRESSION MODELS BASED ON HARVEST OF COOLWATERAND COLDWATER FISHES IN RESERVOIRS

INTRODUCTION

Since 1963, the National Reservoir Research Program (NRRP) has served asa repository for fish standing crop, angler use and harvest, and environmentaldata from United States reservoirs with surface areas of 500 acres or larger.These data are collated, periodically updated, and analyzed to provide broaderinsight into the effects of environmental factors on standing crop and harvestof reservoir fishes. Correlation and multiple regression analysis have beenthe primary analytical techniques used to define important biological-environ­mental relations based on field data. They also have been used to developsimple mathematical expressions to predict reservoir standing crop and harvest,including an assessment of precision or reliability (Jenkins 1967; Jenkins andMorais 1971; Jenkins 1974, 1976, 1977, and 1982). These techniques meshclosely with the objectives of the Habitat Evaluation Procedures Group of theU.S. Fish and Wildlife Service's Western Energy and Land Use Team, which areto provide biologists with methods to more accurately quantify changes in fishand wildlife habitat associated with changes in land and water use.

In a previous analysis, Aggus and Morais (1979) developed regressionformulas to predict standing crops of fishes in warmwater impoundments andcumulative frequency distribution plots to equate these predicted standingcrops to Habitat Suitability Indices (HSI's), as required in the 1978 draftversion of Habitat Evaluation Procedures (HEP) (U.S. Fish and Wildlife Service1978). The standi ng crop estimates were based on cove rotenone data. Thecove rotenone technique is not effective in reservoirs where summer watertemperatures rema in below 24-25°C; therefore, many coo1water and co1dwaterfishes were not included in that analysis. Estimates of angler harvest offerthe largest source of stock assessment data in those impoundments.

The terms "coolwater fishes" and "coldwater fishes" are not clearlydefined. For the purposes of this report, coolwater fishes are designated asmembers of the perch (Percidae) and pike (Esocidae) families, whereas coldwaterforms are members of the trout and salmon family (Salmonidae).

Information is provided for coho salmon (Oncorhynchus kisutch), kokaneesa lmon (Oncorchynchus nerka), cutthroat trout (Sa 1mo clarki), ra i nbow trout(Salmo gairdneri), brown trout (Salmo trutta~rook trout (Salvelinusfontinalis), lake trout (Salvelinus namaycush), northern pike (Esox lucius),muskellunge (Esox masquinongy), yellow perch (Perca flavescens), sauger(Stizostedion canadense), and walleye (Stizostedion vitreum vitreum).

1

HABITAT SUITABLITY INDEX MODELS

Model Applicability

The objective of this report is to develop HSI models for selected cool­water and coldwater fishes in reservoirs. The models, based on estimates ofharvest, are designed for use in existing and proposed coolwater and coldwaterreservoirs in the contiguous United States.

Selection of Variables

Harvest of fish. The Habitat Evaluation Procedures (U.S. Fish and Wild­life Service 1980) utilize two types of information to describe the suitabilityof aquatic habitats for animal species. One type includes direct estimates ofproduction or abundance, whereas the other is based on indicators of habitatsuitability and uses indirect estimates. The techniques employed by the NRRPuse biomass of fish as direct measures of abundance. The use of harvest datato evaluate coolwater and coldwater reservoirs includes only the larger fishin a population. Use of this information to assess habitat suitability,therefore, requires the assumption that suitability of a reservoir for aparticular fish species or species group will be reflected in the biomass offish harvested.

Estimates of harvest used in this analysis are from studies conductedmostly by State fishery management agencies. These include estimates madeover a 30+ year period that incorporate numerous sampling designs. No attemptwas made to adjust for differences in creel census design. Estimates whichincluded only a portion of the angler season or a fraction of the total anglereffort were not used. Data, whenever possible, were analyzed at the specieslevel; however, some f i shery management agencies group fishes into broadertaxonomic categories for reporting. In the case of the coolwater and coldwaterfishes, these included such taxonomic groups as ESBx spp. and "t rout". Thefish species and species groups included in this analysis are listed inTable 1. Agencies collecting the environmental data used for this analysisroutinely report data in English units. To simplify use of the methodology,environmental data are reported in the most commonly measured unit (usuallyEngl ish). Harvest estimates are in metric units to conform with Office ofBiological Services standards.

Environmental variables. The Habitat Evaluation Procedures (U.S. Fishand Wildlife Service 1980) require the assumption that habitat value for fishand wildlife can be determined relative to certain measurable characteristicsof the physical environment. The Procedures also are based on the assumptionthat suitability of a habitat can be numerically characterized by the degreeto which certain life requisites are met, as discussed by Schamberger andFarmer (1978).

In selecting environmental variables for reservoirs, we specified that avariable must: (1) be determined in the early phases of reservoir planningand designing; (2) be easily measured or affected by the design or operationof a reservoir; and (3) have some biological basis; for example, be related tothe behavior or production of one or more reservoir fish species.

2

Since the NRRP began collating and analyzing environmental and fishstanding crop and harvest data, many environmental variables meeting the abovecriteria have been tested. Those listed in Table 1 were selected after removalof highly correlated variables. This was accomplished by first computing asimple correlation matrix for environmental and harvest variables, then retain­ing the environmental variable having the highest correlation to harvestvariables. The removal of highly correlated environmental variables does notgreatly reduce the power of a predictive expression, but does reduce thenumber of measurements required to use the procedure. Variables, such as shoredevelopment, surface area, mean depth, outlet depth, and water-level fluctua­tion, usually are quantified during the early stages of planning for reservoirconstruction. Data on these variables can be obtained from planning andoperating agencies. Water Resources Data, Surface Water and Water QualityRecords 1 provides information needed to compute values, such as storage ratioand dissolved solids. Definitions of environmental variables used in thisreport are presented in Appendix A.

The data base. For this analysis, the NRRP's reservoir data banks weresearched to identify available information on the harvest of coolwater andco 1dwater fi shes in reservoi rs. Li sts of reservoi rs and years of harvestrecords were separated by State and rna i 1ed to appropri ate State fi sheryagencies. Requests for verification and additional harvest or environmentaldata were attached.

Upon return of these requests, new data were 1i sted for computer entryand verifi ed. We further separated data for col dwater fi sheri es into twogroups and analyzed these separately: (1) reservoi rs wi th true coldwaterfisheries; and (2) reservoirs with seasonal (winter) or two-story (thermallystratified) fisheries. This effort yielded environmental and harvest datafrom 57 reservoi rs wi th predomi nant ly coldwater fi sheri es and 39 reservoi rswhere coldwater fisheries existed seasonally or beneath a warmwater (two-story)fishery. In addition, 119 reservoirs with coolwater fisheries were identified.The sample included 421 creel years of records for coolwater fisheries, 172creel years of records for coldwater fisheries, and 110 creel years of recordsfor seasonal or two-story fisheries (Appendix B, Tables B-1 to B-3).

Records of fish harvest ranged from 1-25 years for various reservoirs.We used a mean value for each reservoir as the sampling unit to ensure equalweight for data analysis. These values are summarized in Appendix B,Tables B-1 to B-3.

Data Analysis

The analysis of reservoir harvest data consisted of the following proce­dures: (1) development of multiple regression equations to predict harvest ofselected fishes; (2) conversion of these values to indices of habitat suit­ability; and (3) exploration of discriminant analysis as an alternate quan­titative approach.

iAn annual cooperative publication of the U.S. Geological Survey and Statewater resource agencies from 1964 to the present.

3

Table 1. Fish harvest (dependent) variables and environmental(independent) variables used to develop habitat suitabilityindices for coolwater and coldwater fishes in reservoirs.

Fish harvest variables

Coolwater fishes

Walleye

Sauger

Yellow perch

Northern pike

Muskellunge

aEsox spp.

Coldwater fishes

Rainbow trout

Brook trout

Brown trout

Cutthroat trout

Lake trout

All troutb

Coho salmon

Kokanee salmon

Environmental variables

Surface area

Surface elevation

Mean depth

Outlet depth

Water level fluctuation

Shore development

Storage ratio

Growing season

Total dissolved solids

aIncludes all members of the genus Esox.

bIncludes rainbow, brook, brown, and cutthroat trout.

4

Multiple regression analysis. Multiple regression analysis was used toquantify relationships between environmental (independent) variables andharvest of selected fishes (dependent variables) in reservoirs with coolwater,coldwater, and seasonal or two story fisheries. In this procedure, variationin environmental factors is used to explain variation in harvest of fish. Thepredictive value of any environmental factor is determined by how well a unitchange in that factor is related to a unit change in harvest of a particularspecies. It is assumed that the environmental variables that provide thegreatest predictive value are biologically important, although high correlationdoes not necessarily imply a cause and effect relationship.

Multiple regression models relating environmental variables and harvestof selected fishes were developed from the SAS-79 stepwise method of maximumR2 improvement (Barr et al. 1979). This procedure selects independent vari­ables from all available combinations on the basis of maximum R2 improvementfor each additional variable added in a multiple regression model. We usedboth arithmetic and log values for each dependent and independent variable,thereby considering possible nonlinear relationships.

Regression equations were limited to a maximum of three environmental(independent) variables. Selection of an equation was based on a combinationof the R2 improvement for each independent variable added and the combinedF-values of the regression model. Recommended regression equations for eachspecies or species group in each reservoir type are presented in Appendix C.Equations containing logarithmic expressions of harvest have been adjusted tocorrect geometric means to arithmetic equivalents using the method of Ricker(1975).

Cumulative frequency plots. The Habitat Evaluation Procedures utilizeHabitat Suitability Indices (HSI's) to describe relations between biologicaland environmental features. In this approach, the habitat measures are scaledfrom 0-1, with the most suitable habitats having the highest values. To usethe harvest predictions developed from multiple regression equations in theHabitat Evaluation Procedures, the predictions must be transformed to the 0-1scale used for the HSI. Such a transformation does nothing to improve thepredictive power of the original regression model and does not aid in inter­preting model output.

A predicted harvest can be converted to a 0-1 HSI scale by dividing thepredi ct i on by a harvest va1ue that serves as standard of compari son (U. S.Fish and Wildlife 1980). These values can be known maximum estimates like theones in Tables 2-4. Model outputs derived with this type of model should meetthe HEP assumption that the HSI be linearly related to carrying capacity.Unfortunately, conversion of predicted harvest estimates to an HSI by dividingthe harvest data by the maximum harvest value results in the majority of theHSI's being below 0.5 because the maximum harvest (that is, the standard ofcomparison) is usually much higher than most predicted or observed harvests.This phenomenon does not affect the validity of compensation recomme~dations

deri ved wi th HEP. However, potential HSI model users may be reluctant toaccept a model that rates some of the best available habitat below 0.5 on ascale of a to 1.0.

5

Table 2. Number of observations, means, ranges, and standarddeviations for selected environmental and sport fish harvestvariables from reservoirs with coolwater fisheries.

Number ofVariable observations

IndependentSurface area (acres) 119

Surface elevation (ft) 119

Storage ratio (years) 118

Mean depth (ft) 119

Outlet depth (ft) 119

Fluctuation (ft) 119

Dissolved solids (ppm) 119

Growing season (days) 119

Shore development 119

DependentWalleye harvest (kg/ha) 90

Sauger harvest (kg/ha) 12

Yellow perch harvest(kg/ha) 37

Northern pike harvest(kg/ha) 38

Muskellunge harvest (kg/ha) 13

Esox spp. harvest (kg/ha) 52

aT = < 0.005

6

Mean

14922

1501

0.94

30.7

48.4

15.1

316

174.9

8.98

1. 44

0.26

3.76

3.26

0.39

2.38

Range

510 - 212000

14 - 7665

0.01 - 6.56

4.0 - 250.0

0.0 - 220.0

1.0 - 100.0

10 -1000.0

72 - 330

1. 5 - 38.0

Ta - 26.00

Ta - 1. 54

Ta - 35.09

0.01 - 34.86

0.01 - 0.90

0.01 - 34.86

Standarddeviation

28502

1393

1.29

29.6

44.3

17.3

323

39.7

7.40

3.15

0.40

8.06

7.22

0.35

6.00

Table 3. Number of observations, means, ranges, and standarddeviations for selected environmental and sport fish harvestvariables from reservoirs with predominantly coldwater fisheries.

Number ofVariable observations

IndependentSurface area (acres) 57

Surface elevation (ft) 57

Storage ratio (years) 57

Mean depth (ft) 57

Outlet depth (ft) 57

Fluctuation (ft) 57

Dissolved solids (ppm) 57

Growing season (days) 57

Shore development 57

DependentAll trout harvest (kg/ha) 57

Rainbow trout harvest(kg/ha) 51

Brook trout harvest (kg/ha) 9

Brown trout harvest (kg/ha) 25

Cutthroat trout harvest 15( kg/ha)

Mean

6023

5302

1.23

63.8

90.6

34.6

190

123

3.93

16.72

17.03

0.44

1.06

1.46

Range

530 - 29500

235 - 9869

0.01 - 15.00

4.6 - 228.0

1 - 250

o - 100

16 - 640

6 - 280

1.1 - 15.2

0.05 - 150.89

0.22 - 150.89

Ta - 1.69

0.01 - 5.38

Ta - 6.39

Standarddeviation

7377

2707

2.13

50.0

70.7

26.8

146

51

3.10

25.47

26.49

0.66

2.14

2.07

Lake trout harvest (kg/ha) 7

Coho salmon harvest (kg/ha) 10

Kokanee salmon harvest(kg/ha)

aT = < 0.005

19

7

0.35

1. 78

2.57

0.01 - 0.90

0.04 - 6.92

0.03 - 12.77

0.40

2.14

3.45

Table 4. Number of observations, means. ranges. and standarddeviations for selected environmental and sport fish harvestvariables from reservoirs with seasonal or two-story coldwaterfisheries.

Number of StandardVariable observations Mean Range deviation

IndependentSurface area (acres) 39 20137 500 - 115000 27495

Surface elevation (ft) 39 1399 273 - 5550 1295

Storage ratio (years) 39 1.32 0.03 - 6.56 1.59

Mean depth (ft) 39 56.6 40 - 250 51. 9

Outlet depth (ft) 39 83.6 1 - 220 57.6

Fluctuation (ft) 39 27.1 o - 100 27.3

Dissolved solids (mg/l) 39 246 10 - 1140 288

Growing season (days) 39 186 72- 300 48

Shore development 39 10.31 1.5- 30.0 9.06

DependentRainbow trout harvest 28 2.01 0.01 - 18.32 4.04

( kg/ha)

All trout harvest ( kg/ha) 39 1.61 0.01 - 18.32 3.52

8

We developed Cumulative Frequency (CF) plots of harvest estimates to moreequitably approximate HSI values required in HEP procedures. Harvest estimatesfor a given species or species group were ranked in increasing order. Thecumulative frequency associated with a given harvest value was calculated asthe proportion of harvest estimates that were less than or equal to thatvalue. Cumulative frequency values for predicted harvests can, thus, varybetween 0 and 1. Initial plotting of data for coolwater and coldwater fishesrevealed that frequency distributions of harvest estimates were stronglyskewed, with many small values and few large ones. Harvest estimates were,therefore, plotted on a log scale to provide greater resolution for the manysmall values. An example of the CF plot for walleye is shown in Figure 1.Additional CF plots for species represented in coolwater fisheries, coldwaterfisheries, and seasonal two-story fisheries are presented in Appendix 0,Figures 0-1 to 0-17.

The CF plots include all estimates of harvest available to the NationalReservoir Research Program. These frequently exceed the number of observationsused for multiple regression analysis because environmental variables were notavailable from all reservoirs where harvest estimates were made. It is assumedthat the CF plots represent the true range and frequency of harvest forselected fishes. Obviously, this assumption becomes more valid with a largerand more diverse data source.

Discriminant analysis. The feasibility of classifying reservoirs on thebasis of grouped harvest of selected sport fishes, then predicting groupmembership based on environmental characteristics of a reservoir, was testedwith discriminant analysis. In evaluating this technique, we selected themost widely distributed coolwater fish (walleye) and coldwater fish (rainbowtrout) and divided the reservoirs where each species occurred into four groups,based on estimated annual harvest of these fi shes. Reservoi r groups wereidentified as poor, fair, good, or excellent, based on the ranges of harvestvalues shown in Table 5.

9

CUMULRTIVE FRESUENCY

151 151 151 151 151

~ ..... .J:

151.an

151

D"I

151

ED

151

LD

51::a<r1U1-f

C."

Z:IIr­r­r1-<r1

I. III +--+--+---41---+--4----+-,__+---+--1--+

1.112

I.II~

1.111

1.121

I.I~I

I. III

1.211

I.~II

I .111 +---------+0&..2.111

~.III

11.111

21.111

Figure 1. Cumulative frequency plot of walleye harvest, showing aprocedure for obtaining CF values.

10

Table 5. Classification of walleye and rainbow trout reservoirs,based on observed groupings of estimated annual harvest.

Walleye harvest Rainbow trout harvestClass Minimum Maximum Sample Minimum Maximum Sample

( kg/hal ( kg/hal size ( kg/hal ( kg/hal size

Poor 0.0 0.1 28 0.0 2.2 35

Fair > 0.1 1.1 30 > 2.2 11.2 21

Good > 1.1 2.2 18 > 11.2 22.4 9

Excellent > 2.2 14 > 22.4 12

Discriminant analysis was performed on each data set, using the programof Barr et al. (1979). Covariance matrices were unequal in both instances,and a multivariate normalization procedure was used to equalize the varianceswithin matrices. Each data set was then reanalyzed, using the transformedvalues and the BMDP discriminant function procedure (Brown 1977). The classesdeveloped for harvest were poorly defined by the environmental characteristicsin both tests. Testing of additional species was discontinued pending devel­opment of alternative classification schemes for harvest.

Interpreting Model Outputs

Comput i ng habi tat sui tabll i ty by the procedure descri bed in thi s reportinvolves two steps: (1) obtaining an estimate of harvest from observed dataor the appropriate regression formula; and (2) converting the harvest estimateto a Habitat Suitabilty Index (HSI) using the appropriate maximum harvest orCF plot. Ideally, the user should estimate annual harvest from publishedsources, or from the appropri ate State fi shery management agency, and thencompute the HSI value from the appropriate CF plot or maximum harvest value.

To use the CF plots, first select the appropriate CF plot for a speciesor group of species and reservoir type (coolwater, coldwater, seasonal, ortwo-story coldwater). The calculated or observed value for harvest should belocated on the X-axis, and a perpendicular line extended to the point ofintersection on the CF line. By extending a line parallel to the X-axis fromthis point of intersection to the Y-axis, a 0-1.0 value is obtained (Fig. 1).This value represents the proportion of available harvest estimates that wereless than, or equal to, the observed or predicted harvest and can be used asan HSI.

In planned reservoirs or impoundments where harvest data are not avail­able, a predicted harvest can be obtained from the appropriate multiple regres­sion equation in Appendix C. Use of these equations requires estimates of the

11

environmental parameters defined in Appendix A. Reservoir planning or operat­ing agencies can usually supply needed data. Water Resources Data, publishedfor each State by the U.S. Geological Survey, is also a good data source. Theregression equations can be solved easily with a desktop or pocket calculatorand standard log tables. Harvest estimates in logarithmic form require thatthe antilog be obtained before proceeding with the evaluation. Predictedharvest values can then be converted to an HSI by dividing by the appropriatemaximum value for harvest provided in Tables 2-4.

A regression model will not reflect a real world situation, but insightsinto the strengths of the relations can be gained by examining the followingrelationships: (1) the number of observations used to develop each equation;(2) the coefficient of multiple determination (Rt) (printed below each regres­sion equation), which defines the percentage of variability in harvestexplained by the equation; and (3) the probability (P), which is the chance ofobtaining an equal or larger F-value when the hypothesis of no correlation istrue. Ranges and standard deviations of harvest and environmental variablesfor reservoirs included in each sample are provided in Tables 2-4. Thesestatistics prov tde information that aids in determining whe-ther or not themodel should be applied. Environmental parameters should fall within therange of values used to develop the regression model or the model should notbe used.

Where computing facllities are available, the multiple regression equa­tions and CF plots can be programmed to permit users to input environmentaldata and receive harvest and cumulative frequency values as output. Themultiple regression equations presented herein are easily programmed for thispurpose. However, CF plots present a more complicated programming problem inthat cumulative frequency distributions of harvest are variable, but tend tobe skewed with few relatively high values and with means well above the median.A cumulative Weibull distribution provides reliable approximation of the CFdistribution. Characteristics of the Weibull distribution and computationalprocedures are described in Johnson and Kotz (1970).

RECOMMENDATIONS

The value of the Habitat Evaluation Procedures rests largely on how wellHSI models used in the Procedures predict environmental impacts for specifichabitats and the degree of acceptance by persons charged with using theProcedures. The NRRP's multiple regression procedures for predicting fishstanding crop and harvest have been widely used within the fisheries community.However, harvest data are an i ncomp 1ete method of populat i on assessment.Conceptually, harvest should be a satisfactory index of habitat suitability,because it represents the desired end product for resource managers and isultimately influenced by carrying capacity. In most instances, the environmen­tal factors tested accounted for less than 50% of the variation in harvest ofselected fishes. However, it should be recognized that the use of environmen­tal factors to predict harvest in coolwater and coldwater reservoirs mayremain relatively imprecise. Some coolwater fishes and a large percentage ofcoldwater fishes are maintained in reservoirs through put-and-take or put-grow­and-take fisheries. Factors, such as stocking rate or access by fishermen, may

12

be equally as important as environmental variables in predicting harvest ofcoolwater and coldwater fishes. In addition, harvest estimates are not exactand may be influenced by differences in creel census designs.

Discriminant analysis did not produce precise classification schemes inreservoirs, based on harvest of walleye and rainbow trout. However, evaluationof the technique should continue. Harvest data in this study were grouped byarrangi ng observations in ascendi ng order and arbitrari ly separating theseinto levels of harvest, based on pulses in the distribution of values. Quanti­tative methods for identifying appropriate levels of harvest should be soughtand tested before the procedure is abandoned.

REFERENCES

Aggus, L. R., and D. 1. Morais. 1979. Habitat suitability index equationsfor reservoirs based on standing crop of fish. U.S. Dept. Int., FishWildl. Serv., Off. Biol. Serv., Ft. Collins, CO. 120 pp. mimeo.

Barr, A. J., J. H. Goodnight, J. P. Sall, and J. T. Helwig. 1979. A usersguide to SAS. SAS Institute Inc., Sparks Press, Raleigh, NC. 329 pp.

Brown, M. B. (ed.) 1977. BMDP 77, biomedical computer programs, P-Series.University of California Press, Los Angeles.

Jenkins, R. M. 1967. The influence of some environmental factors on standingcrop and harvest of fishes in U.S. reservoirs. Pages 298-321 in ReservoirFishery Resources Symposium, Southern Div. Am. Fish. Soc., Athens, GA.

1974. Reservoir management prognosis: migraines or miracles. Proc.Southeastern Assoc. Game and Fish Commissioners 27:374-385.

1976. Prediction of fish production in Oklahoma reservoirs on thebasis of environmental variables. Ann. Okla. Acad. Sci. 5:11-20.

1977. Prediction of fish biomass, harvest, and prey-predator rela­tions in reservoirs. Pages 371-384 in W. Van Winkle (ed.), Proceedingsof the Conference on Assessing the Effects of Power-Plant-InducedMortality on Fish Populations. Pergamon Press, NY.

. 1982. The morpoedaphic index and reservoir fish production. Trans.----:--

Am. Fish. Soc. 111(2):133-140.

Jenkins, R. M., and D. 1. Morais. 1971. Reservoir sport fishing effort andharvest in relation to environmental variables. Pages 371-384 inG.E. Hall (ed.), Reservoir fisheries and limnology. Am. Fish. Soc. SpeC":"Publ.8. Washington, DC.

Johnson, N. L., and S. Kotz. 1970. Continuous univariate distributions - 1.John Wiley &Sons, NY. 300 pp.

13

Ricker» W. E. 1975. Computation and interpretation of biological statisticsof fish populations. Fish. Res. Board Canada» Bull. 191. 382 pp.

Schamberger» M.» and A. Farmer. 1978. The habitat evaluation procedures:their appl ication in project planning and impact evaluation. Trans. N.Am. Wildl. Nat. Resour. Conf. 43:274-283.

U.S. Fish and Wildlife Service. 1978. Habitat Evaluation Procedures. U.S.Dept. Int., Fish Wildl. Servo Fort Collins, CO. 66 pp + App. (Draftreport) .

U.S.Fish and Wildlife Service. 1980. Habitat Evaluation Procedures (HEP).ESMI02. U.S. Dept Int., Fish Wildl. Servo Washington, DC. n.p.

14

APPENDIX A. DEFINITIONS OF ENVIRONMENTAL VARIABLES USED IN EQUATIONSFOR PREDICTING HARVEST OF COOLWATER AND COLDWATER FISHES IN RESERVOIRS

surface area - In acres at average annual pool level or use conservation pool,operating pool, summer pool, or power pool area, as listed by operatingagency.

surface elevation - In feet above mean sea level at listed surface area.

volume - In acre feet at the elevation listed for surface area.

mean depth - Average depth in feet (volume divided by surface area).

outlet depth - Midline depth of principal outlet, in feet below listed eleva­tion.

water level fluctuation - Mean annual vertical fluctuation of reservoir surfacelevel, in feet.

shoreline length - In miles at listed surface area.

shore development - The ratio of shoreline length to the circumference of acircle equal in area to that of the reservoir.

storage ratio - Reservoir water volume in acre-feet (at listed surface eleva­tion) divided by the average annual water release (discharge) in acrefeet.

growing season - Average number of days between last and first frost annually.

total dissolved solids - Residue after evaporation at 1800 C in ppm.

15

APPENDIX B. LIST OF RESERVOIRS AND YEARS OF RECORD USED TO DEVELOPPREDICTORS OF HARVEST IN RESERVOIRS WITH COOLWATER, COLDWATER, ANDSEASONAL OR TWO-STORY COLDWATER FISHERIES

Table B-1. Reservoir names and numbers of years of harvest records usedto develop suitability indices for coolwater fishes in reservoirs.

Reservoirand State

Years ofrecords

Reservoirand State

Years ofrecords

Altus, OK 1 Douglas, TN 1Anderson Ranch, ID 10 East Lynn, WV 3Angostura, SD 2 Fall River, KS 3Arbuckle, OK 1 Fletcher Pond, MI 1AuTrain, MI 1 Fort Cobb, OK 2Banks, WA 1 Fort Peck, MT 2Belleville, MI 1 Fort Supp ly, OK 1Bi g Creek, IA 4 Foss, OK 1Blue Ridge, GA 2 Francis Case, SD 3Boyd, CO 9 Glen Elder, KS 3Boysen, WY 2 Greenwood, SC 4Buckhorn, KY 6 Hamlin, MI 1Cagl es Mi 11 , IN 2 Hartwell, SC 4Canton, OK 5 Hefner, OK 1Canyon, AZ 8 Hodenpyle and Ti ppy , MI 1Ca rry Fa 11 s , NY 1 Ho 11 oway, MI 1Cascade, ID 8 Hubbard, MI 1Cave Run, KY 2 Hugh Butler, NE 1Cedar Bluff, KS 3 Hugo, OK 1Center Hill, TN 7 Jackson, GA 4Cheney, KS 3 Jocassee, SC 3Cherry Creek, CO 1 John Redmond, KS 3Chickamauga, TN 2 Kanopolis, KS 9Chickasha, OK 1 Kent, MI 1Chippewa Flowage, WI 1 Kentucky, KY 6Clark Hill, GA 8 Keowee, SC 5Claytor, VA 1 Keyhole, WY 5Cohoon, VA 6 Keystone, OK 3Conchas, NM 1 Kirwin, KS 3Coralville, IA 1 Lake of the Ozarks, MO 16Council Grove, KS 3 Lovewell, KS 3Dale Hollow, TN 10 Macbr ide, IA 2Deep Creek, MD 5 Marion, KS 3

16

Table B-1. (concluded).

Reservoirand State

Years ofrecords

Reservoirand State

Years ofrecords

McConaughy, NE 1 Saguaro, AZ 6Meade, VA 6 Seminoe, WY 4Melton Hi 11, TN 3 Sharpe, SO 1Melvern, KS 3 Shelbyville, IL 1Michigamme, MI 1 Smith Mountain, VA 5Mi lford, KS 3 South Holston, TN 4Mina, SO 1 Spring, IL 3Monroe, IN 1 Stockton, MO 5Nickajack, TN 1 Stonecoal, WV 3Nolin, KY 8 Stony, MI 1Norris, TN 6 Strunk, NE 1Norton, KS 3 Sutton, WV 2Nottely, GA 3 Taneycomo, MO 1101 d Hi ckory, TN 3 Tenkiller Ferry, OK 4Perry, KS 3 Thomas Hi 11, MO 5Pomme Oe Terre, MO 8 Tuttle Creek, KS 3Pomona, KS 3 Upper Lake Mary, AZ 4Pontiac, MI 1 Ute, NM 1Potholes, WA 1 Vallecito, CO 1Powell, UT 3 Watauga, TN 9Prince, VA 6 Watts Bar, TN 1Quabbin, MA 25 Webster, KS 3Rabun, GA 2 Western Branch, VA 4Rathbun, IA 4 Whitney Point, NY 1Reedsburg, MI 1 Wil son, KS 3Robert S. Kerr, OK 1 Woods, TN 6Rough River, KY 8

TOTAL 421

17

Table B-2. Reservoir names and numbers of years of harvest records usedto develop suitability indices for coldwater fishes in reservoirs with pre­dominantly coldwater fisheries.

Reservoir Years of Reservoir Years ofand State records and State records

Alcova, WY 4 Mohave, AZ 3Anderson Ranch, ID 10 Navajo, NM 1Banks, WA 1 Orovill e, CA 2Beards1ey, CA 6 Palisades, ID 5Bi g Lake, AZ 2 Pathfinder, WY 3Big Sandy, WY 1 Pishkun, MT 1Blue Mesa, CO 7 Ri ffl e, CO 1Bluewater, NM 3 Ruedi, CO 1Buffalo Bill, WY 3 Scofield, UT 1Cottage Grove, OR 1 Seminoe, WY 4Crowley, CA 1 Shadow Mountain, CO 2Deer Creek, UT 5 Shasta, CA 2Di 11 on, CO 1 Spaulding, CA 1Dworshak, ID 1 Starvation, UT 5Ennis, MT 1 Steinaker, UT 2Flaming Gorge, UT 13 Storrie, NM 3Fontene11 e, WY 4 Strawberry, UT 9Georgetown, MT 3 Swift, WA 2Gibson, MT 1 Taneycomo, MO 10Glendo, WY 2 Taylor Park, CO 1Granby, CO 4 Turquoise, CO 2Green Mountain, CO 1 Twin Lakes, CO 2Hebgen, MT 1 Upper Lake Ma ry , AZ 4Henrys Lake, ID 5 Va 11 ecito, CO 1Ice House, CA 1 Wi 1d Horse, NV 8John, CO 6 Willow Creek, (Harrison)a MT 1Mackay, ID 1 Willow Creek, (Sun River)a MT 1Mayfield, WA 1 Yale, WA 2Merwin (Ariel)a, WA 2

TOTAL 172

aSynonym for reservoir name.

18

Table B-3. Reservoir names and numbers of years of harvest records usedto develop suitability indices for coldwater fishes in reservoirs withseasonal or two-story coldwater fisheries.

Reservoirand State

Years ofrecords

Reservoirand State

Years ofrecords

Angostura, SO 1 McConaughy, NE 1AuTrain, MT 2 Mead, AZ 1Be 11 evill e, MD 1 Me lton Hi 11, TN 2Blue Ridge, GA 1 Merle Collins, CA 7Broken Bow, OK 1 Old Hickory, TN 1Bull Shoals, AR 4 Pine Flat, CA 4Cachuma, CA 1 Piru, CA 1Cascade, ID 8 Potholes, WA 1Cherry Creek, CO 1 Powell, UT 3Clark Hill, GA 2 Quabbin, MA 25Dale Hollow, TN 5 Robert S. Kerr, OK 1Fish Trap, KY 1 Round Valley, Nj 1Fol som, CA 1 Smith Mountain, VA 1Ha rtwe11, SC 4 South Helson, TN 4Ho 11 oway, MI 1 Spruce Run, NJ 2Hubbard, MI 1 Stony, MI 1Isabella, CA 2 Sutton, WV 1Jocassee, SC 3 Table Rock, MO 1Key Hole, WY 1 Watauga, TN 9Laurel, KY 2

TOTAL 110

19

APPENDIX C. MULTIPLE REGRESSION FORMULAS FOR ESTIMATING HARVEST(KG/HA) OF COOLWATER AND COLDWATER FISHES FROM RESERVOIRS.

The coefficient of multiple determination (R2) provldes an approximation

of the percentage of variation in harvest explained by the equation. Theprobability (P) indicates the chance of obtaining an IIF II value as large, orlarger. when the hypothesis of no correlation is true.

Reservoirs which had incomplete or missing environmental variables werenot used in the final regression equations. Therefore, the number of observa­t ions in the equations below do not necessarily correspond to the number ofobservations listed in Tables B-1 to B-3 or in cumulative frequency plots(Figs. 0-1 to 0-16).

Regression equations were omitted for sauger and brown trout, becausestatistically significant correlations were not found between environmentalvariables and their respective harvests.

COOLWATER FISHERIES

Walleye

Log (harvest of walleye) = 2.1299 + 0.3563 log (storage ratio) - 0.8364 log(shore development) - 0.00622 (growing season).

Number of observations =89

Yellow Perch

R2 =0.315 P = 0.0001

Log (harvest of yellow perch) = 3.7117 - 0.0142 (growing season) - 0.7530log (outlet depth).

Number of observations = 37

Northern Pik.e

R2 = 0.378 P = 0.0004

Log (harvest of northern pik.e) = 3.7882 - 0.0177 (growing season) - 0.8447log (outlet depth).

Number of observations = 37 R2 = 0.665

20

P = 0.0001

Muske11 unge

Log (harvest of muskellunge) = 2.8421 - 0.0117 (growing season) - 0.9585log (water level fluctuation).

Number of observations = 12 R2 = 0.518 P = 0.0370

Esox spp.

Log (harvest of Esox spp.) = 3.4687 - 0.0142 (growing season) - 0.00108(total disso1ved-soTids) - 0.7573 log (outlet depth).

Number of observations = 52

COLDWATER FISHERIES

All Trout

R2 = 0.634 P = 0.0001

Harvest of all trout = 137.55 - 26.94 log (surface area) - 30.68 log(mean depth) + 50.13 log (shore development).

Number of observations = 56

Rainbow Trout

R2 = 0.359 P = 0.0001

Harvest of rainbow trout = 148.79 - 31.53 log (surface area) - 28.16 log(mean depth) + 51.38 log (shore development).

Number of observations = 50

Brook Trout

R2 = 0.372 P = 0.0001

Harvest of brook trout = 2.449 + 0.0354 (water level fluctuation) - 2.337log (water level fluctuation).

Number of observations = 9

Cutthroat Trout

R2 = 0.794 P = 0.0087

Log (harvest of cutthroat trout) = 4.8065 - 3.0508 log (mean depth) + 0.0103(outlet depth).

Number of observations = 15 R2 =0.487

21

P = 0.0255

Coho Salmon

Log (harvest of coho salmon) = 1.0102 log - 2.4191 log (shore development)+ 0.0149 (water level fluctuation).

Number of observations = 9

Kokanee Salmon

R2 = 0.755 P = 0.0147

Harvest of kokanee salmon =8.771 + 2.682 (storage ratio) - 4.682 log(growing season).

Number of observations = 19

Lake Trout

R2 = 0.640 P = 0.0003

Harvest of lake trout = 1.826 - 0.8714 log (outlet depth)

Number of observations = 6 R2 =0.950 P =0.0012

SEASONAL AND TWO-STORY COLDWATER FISHERIES

All Trout

Harvest of all trout = -7.765 + 0.00194 (surface elevation) - 0.03534(mean depth) + 0.0465 (growing season).

Number of observations = 39

Rainbow Trout

R2 = 0.452 P = 0.0001

Log (harvest of rainbow trout) = 1.4068 - 0.5862 log (surface area) + 0.3122log (storage ratio) + 0.7855 log (water level fluctuation).

Number of observations =28 R2 = 0.551

22

P = 0.0003

APPENDIX D. CUMULATIVE FREQUENCY PLOTS OF HARVEST OF SELECTED SPORTFISHES IN RESERVOIRS WITH COOLWATER, COLDWATER, AND SEASONAL OR

TWO-STORY COLDWATER FISHERIES

CUMULATIVE FRESUENCY

151 151 151

hJ

151 151 151 151 151 151 151. ...... .c VI a'I III La

:z::D::II<PIU1....C..,:E::DrrPI-<I"'l

S. SSI +--.+-+--.+-+--+-+--+-+--+-+

S.SS2

S.SSS

S.SIS

S.S2S

S.SSS

S.ISS

S.2SS

S.SSS

I.SSS

2.SSS

S.SSS

IS.SSS

2S.SSS

Figure 0-1. Cumulative frequency plot of walleye harvest inreservoirs with coolwater fisheries.

23

(UMULATIVE FRE8UEN(Y

151

h..I.LaJ

151

..J:

151

VI

151

m.-..I

151

a::I

151

LD

".""S +--+--+---I--t--~-+--+--+--+---+

"."'""."2"

%:a::a<fT1111.... "."s"C."

U1:a

a, '"''t::II1fT1::a.... ".2""~II1<,%:a'oJ

".s""

I . """

2. """ +--+-~----i~-+--+---+-+--+-~----i

Figure 0-2. Cumulative frequency plot of sauger harvest inreservoirs with coolwater fisheries.

24

CUMULRTIVE tREBUENCY

m m m m m m m m m mhJ lIJ .J: VI rn ...., m La

B. BB2 ..J.---I--4-----If---f--+---+-+--+-_+_--+

~

::D::a<JT1\l1-t

C."

-<JT1rrcx:"tIJT1::a,....~

B.BBS

B.BIB

B.B2B

B.BSB

B.IBB

B.2BB

B.SBB

I.BBB

2.BBB

S.BBB

IB.BBB

2B.BBB

Figure D-3. Cumulative frequency plot of yellow perch harvest inreservoirs with coolwater fisheres.

25

(UHULRTIV£ rR£BU£N(Y

151 151 151 151

I'Ll 1IJ

151 151 151..J: V1 D'1

151 151 151

..... D:I Ul

B.BB~ +--+--+---+-~--lI--.......-t---t--.......-1'

B.BIB

B.B2B:z::rJ::a

B.B~B<1""1U1~

B. IBBc...,z B.2BBc::a~:z:1""1 B.SilB::az-a I .IBB-~1""1

"" 2.Iill~D"I......:z: S.ilBB::a"J

IB.BBB

21.IBB

Figure D-4. Cumulative frequency plot of northern pike harvest inreservoirs with coolwater fisheries.

26

(UHULRTIVE rRE~UEN(Y

m m m m m m m m m mhJ LaJ .J: &'1 En .... m La

a.a Ia +--+--+--+--+--+--+--+--+--+---t

llI.ra21l1

:::t::a::II<.."IJ1-f

ra.ras:rac'"'l

XCIJ1~ Ill. I aa.."r-r-czI:i'l.."- ra.2ralll~I:i'l<,:::t::a.....

ra.S:1lI1l1

I . raralll +---+--+--+--+--+--+--+--+--+----+

Figure D-5. Cumulative frequency plot of muskellunge harvest inreservoirs with coolwater fisheries.

27

CUMULATIVE rRESUENCY

151 151 151 151 151 151

h.I .... oJ: VI

151 151 151 151.D1 ...., m La

I. mI I +---+--t-----t--+---it---+--t----t---t---t

m.IS:1

% I. III::a::u< 1.2111"'1I.lI....C..,

8.5:881"'1I.lIC)( 1.188,...~In 2.m18<,:E:::D-

S.81m

Im.1lI88

21.8BIlI

Figure 0-6. Cumulative frequency plot of Esox spp. harvest inreservoirs with coolwater fisheries. ----

28

CUMULRTIVE rRESUENCY

151 151 III.hJ

III III.111 In

III

III

III

u:am.2mm +---+--+---1I---+--f----f--+---+--+---+

m.smm%:D I. mmm::a<r'I111-f 2.mmmc....::a:D- s.mmmzIJICz:-f Im.mmm::acc::-f- 2m.mam~a1

"%:D- sm.mmm

Ima.mam

2mm.mmm

Figure 0-7. Cumulative frequency plot of rainbow trout harvest inreservoirs with coldwater fisheries.

29

CUMULRTIVE rRESUENCY

m s ssm m m m m m. . ..f\J .... .I: VI en .... Ell LII

I. I 1(/1 ......-+-+---+--+---4--+--~.-..+--+--+

1.12B

:c 1.lil:II::II<r1Ln

I. III....c:r."

Ell 1.211::IICIEZ

....::II I.illc:rc....,..

I . BII~en<,:c:II 2.BIB-

i.111

Figure 0-8. Cumulative frequency plot of brown trout harvestin reservoirs with coldwater fisheries.

30

Figure 0-9. Cumulative frequency plot of cutthroat trout harvest inreservoirs with coldwater fisheries.

31

CUMULRTIVE FREQUENCY

m m m m m m m m m mhJ LaJ .J: VI In -...I D:I LQ

1i!.1i!1i!1

1i!.1i!1i!2

::E:1iJ.IiJIi!S

::D::a< 1i!.1i!11i!1"'1tn-I

c 1iJ.1i!21i!..,IR::acc 1i!.Ii!SIi!~

-I::a 1iJ. I Ii!IiJcJ::-I,...

1i!.21i!1i!~In<,::E:::D...... Ii!.SIi!IlI

I . Ii!Ii!Ii!

2. IIIIi!III

Figure D-10. Cumulative frequency plot of brook trout harvest inreservoirs with coldwater fisheries.

32

CUMULRTIVE FREBUENCY

m m m ~ m m m m m mh.I LaJ ..r: V1 D'1 ...,J m LII

m.msm +r---+--I----II---+--+---+--t---t--t----t

a. ram

m.2mm

B.smm

I . mJam

s.mma-~til ia. ama<,~::D..... za. mma

sm.mmm

Imm.mma

2ml2l .mmm+---+--I----II---+--+---+--t---t--t-----t

Figure 0-11. Cumulative frequency plot of total trout harvest inreservoirs with coldwater fisheries.

33

CUMULATIVE FRESUENCY

lSI lSI lSI 151 lSI lSI lSI

1\.1 I.AJ oJ: V1 In

lSI lSI. ..... a:J

lSI

La

111. BIB -r---+--1----.,.....-+-+--+--+---It--..........---1'

B.B2B

:c:::tI:;a<r1U1.... B.B~Bc~

r:::tI

"'"r1 1a.llaB....:act:...."'"' B.2BB~Ii1.......:c:::tI-

B.~BI1I

I . BI1II1I +--+--+---t--+---l~-+--~-+--+--+

Figure D-12. Cumulative frequency plot of lake trout harvest inreservoirs with coldwater fisheries.

34

CUMULRTIVE FREBUENCY

I.B~e

151

\

lSI

hJ

lSI...c

lSI.V1

lSI.D'I

lSI

.....lSI

to

lSI

LD

:E: 1i1. I Ii1Ii1:D:::a<1""'1lit.....c 1i1.2BIi1""1

......C:E:C

lit:D B.5:11i1,...:zcz.....~ I . Ii1Ii1Ii1Ei"I.......:E::D-

2.1i111i1

5: .11i11i1 +-......--+-~I---+--~-+- .......---t--+----t

Figure 0-13. Cumulative frequency plot of coho salmon harvestin reservoirs with coldwater fisheries.

35

B.I2ISB

::J: 121. IBO::D::D<f""IIn B.2121121-t

c""1

~c l2I.som~:czrrIrrI

In I .1211219:cr::J:c

2.m12l9z-~D1<,::J::c S.OI2lI2l-

1121.(111219

2m.amm

CUMULATIVE rRE9UENCY

m m m m m m m m m mhJ ~ .J: V1 D'1 ..... OJ U1

Figure 0-14. Cumulative frequency plot of kokanee salmon harvestin reservoirs with coldwater fisheries.

36

I5J lSI I5J

I'Ll

CUMULRTIVE FREBUENCY

I5J lSI lSI lSI lSI lSI. . . . . .1.&1 .3:. ." at .... lEI

I5J.La

11.11 IIlI +---+--+---ll--........-+--+-+---t--t-----t

1.1121

1.115:1

%:II II. III::II<1"1111-f 1.2111c~

-f::II 1.5:111cc:-f- I .111111~11'1......%

2.111:II-5:.1111

11.1111

21.11111

Figure 0-15. Cumulative frequency plot of total trout harvestin reservoirs with seasonal or two-story fisheries.

37

CUHUI.AT I VE F"RESUEN<Y

m m m m m m m m m mhJ ..... .J: VI C'1 -.I D:I La

a.a 11I1 +--+-+--+--of---f--f---f--...--+--t

1.121

% I.mS:I:a::a<1""1lit I. IJal.....c....::a 1i1.21i11:a-zIRc I.S:1i11:IE:

.....::ac I . IBIc.....-~ 2.mBIEi1"-%:a- s.mml

II.BBI

211.1118

Figure 0-16. Cumulative frequency plot of rainbow trout harvestin reservoirs with seasonal or two-story fisheries.

38

50271-1"'

REPORT ~~MENTATION : 1. RE'F'wsl8BS-82/1 O. 25 i2.i

IJ. Reclllieftl'. Acc..- No.

I~ Title and Subtitle

Habitat Suitability Index Models: Regression Models Based OnHarvest Of Coolwater and Coldwater Fishes In Reservoirs

I 50 RellOlt O.e

\September 1982I.

7. Autllor(.'Larry R. Aggus and William M. Bivin

I I. ""rlarmin. O...nlUClon R--. No.

10. ,,",i_IT.../Work Unit No.

lL C4ntneCfCl or Granc(O) No.

eCl

U.S. Fish and Wildlife ServiceNational Reservoir Research Programr~~----=-~--~~------~I

Fayettev il le , AR

9. Perlarmi". O"."izaClon Nlme and Addr...

(0)

Western Energy and Land Use TeamOffice of Biological ServicesFish and Wildlife ServiceU.S. Department of the InteriorWashington, D.C. 20240

II. ADstreet (Limit: 200 _rd.)--------:------------------------------11This report presents methods designed to permit habitat classification of reservoirs thatcontain coolwater, coldwater, and seasonal two-story fisheries. Multiple regressionequations describing relations between reservoir environmental characteristics andbiomass harvest of selected sport fish species or groups are presented. Cumulativefrequency plots of known harvest estimates from the various classes of reservoirs arepresented to facilitate conversion of harvest predictions to Habitat Suitability Indices(HSI's). Detailed descriptions and limitations of the procedures are discussed.

17. Oocument Anl'Y'i. a. OeacriptOr'l

ReservoirsHabitabil ity

I WaterAquatic biology

FisheriesFishesBiomassFresh water

I b. Identiliers/Ooe"·E"ded Terms

Habitat classification

I

Coolwater fishesColdwater fishesSeasonal two-story fishes

Environmental characteristicsSport fisheries harvestHabitat Suitability Index

e. COSATI ".ld/Group

lL ....vatiaDility Statement

RELEASE UNLIMITED

:S.. ANSI-Z39.18)

'trU.S. GOVERNMENT PRINTING OFFICE:l982-ee0-t23 , 423.

19. Secunty Cl.... (Thi. Reaort)

UNCLASSIFIEDZOo Securoty cr••• (ThIs PI.e'

UNCLASSIFIED

: 21. No. af PI."

:ii - yjij +1Rnn, 22. Price

O",ONAL "ORM m e.a-rn(Formeny NTI5-351OeDarl'me"! of Commerce

o

o

GXD

LEGEND

Headquarters· Office of BiologicalServices, Washington, D.C.National Coastal Ecosystems Team,Slidell, La.Western Energy and Land Use Team,Fort Collins, Co.Regional Offices

REGION 1Regional DirectorU.S. Fish and WildlifeServiceLloyd Five Hundred Building, Suite 1692500 N.E. Multnomah StreetPortland, Oregon 97232

REGION 2Regional DirectorU.S. Fish and Wildlife ServiceP.O. Box 1306Albuquerque, ew Mexico 87103

REGION 3Regional DirectorU.S. Fish and Wildlife Servicefederal Building, Fort SnellingTwin Cities, Minnesota 55111

U.S. FISH A D WILDLIFE SERVICEREGIONAL OFFICES

REGION 4Regional DirectorU.S. Fish and Wildlife ServiceRichard B. Russell Building75 Spring Street, S.W.Atlanta, Georgia 30303

REGION 5Regional Director

.S. FI h and Wildlife ServiceOne Gateway Center

ewton Corner, Mas achusens 02158

REGION 6Regional DirectorU.S. Fish and Wildlife ServiceP.O. Box 25486Denver Federal CenterDenver. Colorado 80225

REGION 7Regional DirectorU.S. Fishand Wildlife SelYice1011 E.Tudor RoadAnchorqe. Alaska 99S03

As the Nation's principal conservation agency, the Department of the Interior has respon­sibility for most of our .nat ionally owned public lands end natural resources. This includesfostering the wisest use of our land and water resources, protecting our fish and wildlife,preserving th&environmental and cultural values of our national parks and historical places.and providing for the enjoyment of life throulh outdoor recreation. The Department 8S­sesses our energy and mineral resources and works to assure that their development is inthe best interests of all our people. The Department also has a major responsibility forAmerican Indian reservation communities and for people who live in island territories underU.S. administration.

.....tiS" AYrIUM .....

~ • .HVH ·• .

DEPARTMENT OF THE INTERIORu.s. FISH AND WILDLIFE SERVICE