Loading...
HomeMy WebLinkAboutAPA2836"'"-=--.'::'"""=------------"----~------------------- -- .... .... ........ ALASKA DEPARTMENT OF FISH AND GAME SUSITNA HYDRO AQUATIC STUDIES REPORT SERIES ,- I I r- I r I ALASKA DEPARTMENT OF FISH AND GAME SUSITNA RIVER AQUATIC STUDIES PROGRAM REPORT NO.7 RESIDENT AND JUVENILE ANADROMOUS FISH INVESTIGATIONS (MAY -OCTOBER 1984) PARTS 1 AND 2 Editors:Dana C.Schmidt,Stephen S.Hale, and Drew L.Crawford Prepared for:Alaska Power Authority 334 W.Fifth Avenue,Second Floor Anchorage,Alaska 99501 July 1985 ARLIS Alaska Resources Library &InformatIOn ServIces Anchorage,Alaska -. I~ -..- 0) o:::t No:::t,c~o:::t 0 0 0 l!) l!) I" ('I) ('I) r-- PREFACE This report is one of a series of reports prepared for the Alaska Power Authority (APA)by the Alaska Department of Fish and Game (ADF&G)to provide information to be used in evaluating the feasib"llity of the proposed Susitna Hydroelectric Project.The ADF&G Susitna River Aquatic Studies Program was initiated in November 1980. The report covers studi es conducted from May through October 1984 of juvenile salmon and resident fish species of the Susitna River.In addition,some information is included on overwintering of resident fish radio-tagged in 1983.The majority of the effort during the 1984 open-water season was on the lower river (from the mouth to the Chulitna River confluence).No studies were conducted this year in the area above Devil Canyon.This volume consists of three parts. Part 1 (RSA Tasks 16A and 16B)covers the migration and growth of juvenile salmon.Coded wire tagging of chum and sockeye fry in the middle river (Chulitna River confluence to Devil Canyon)and collecting of a 11 species of outmi grati ng fry at Tal keetna Station were simil ar to 1983 studies.In addition,a mark-and-recapture cold branding study was conducted in tributaries,sloughs,and side channels of the middle river to obtain estimates of chinook and coho juvenile salmon population size and residence time in these rearing areas.Also,outmigrant traps were operated near the mouth of the ri ver at Fl athorn Station (River Mi 1e 22.4)to obtain a timing index of outmigration from the lower river. A statistical time series analysis of 1983 and 1984 discharge, turbidity,and juvenile salmon outmigration data from the middle river is included as an appendix. Studies of the distribution and relative abundance of juvenile salmon and modelling of rearing habitat in the lower river are discussed in Part 2 (RSA Tasks 14 and 36).These studies were similar to those conducted in the middle river in 1983.Habitat suitability criteria developed for the middle river were used for the lower river unless evidence of different conditions in the lower river necessitated modifi- cations.Results from habitat modelling at 14 RJHAB model sites and 6 IFIM model sites are presented.The RJHAB and IFIM models were compared by using both at two sites.IFIM model calibration is contained in Appendix D. Part 3 (RSA Task 14)presents the results of resident fish studies in both the mi dd1 e and lower river.Moni tori ng of fi sh movement through use of radio tags was continued.Index sites in the middle river were sampled as part of the long term monitoring effort.Population esti- mates for rainbow trout,Arctic grayling,round whitefish,and longnose suckers in the middle river were made from multiple year mark-recapture data. Questions concerning this report should be directed to: Alaska Power Authority 334 West 5th Avenue 'Anchorage,Alaska 99501 Telephone:(907)276-0001 ARLIS Alaska Resources Library &InformatIOn SenrtCf' Anchorage,Alaska". TITLES IN THIS SERIES Report Publication Number Title Date ~1 Adult Anadromous Fish Investigations:April 1984 May -October 1983 ~2 Resident and Juvenile Anadromous Fish July 1984 Investigations:May -October 1983 r-3 Aquatic Habitat and Instream Flow September 1984 Investigations:May -October 1983 4 Access and Transmission Corridor Aquatic September 1984 Investigations:May -October 1983 5 Winter Aquatic Investigations:March 1985 ,rQ...September 1983 to May 1984 6 Adult Anadromous Fish Investigations:June 1985 ,.....May -October 1984 7 Resident and Juvenile Anadromous Fish July 1985 ,-,Investigations:May -October 1984 8 Availability of Invertebrate Food Sources 1985 for Rearing Juvenile Chinook Salmon in Turbid Susitna River Habitats 9 Summary of Salmon Fishery Data for June 1985 N Selected Middle Susitna River Sites ;8 - -- .- Part 1. Pa rt 2. Pa rt 3. CONTENTS OF REPORT NO.7 The Migration and Growth of Juvenile Salmon in the Sus itna Ri ver. The Relative Abundance,Distribution,and Instream Flow Relationships of Juvenile Salmon in the Lower Susitna Ri ver. Resident Fish Distribution and Life History in the Susitna River below Devil Canyon. ...... - - - .- - PART 1 The Migration and Growth of Juvenile Salmon in the Susitna River - ,..,. -. - THE MIGRATION AND GROWTH OF JUVENILE SALMON IN THE SUSITNA RIVER Report No.7,Part 1 by Kent J.Roth and Michael E.Stratton Alaska Department of Fish and Game Susitna River Aquatic Studies Progra~ 620 East 10th Avenue,Suite 302 Anchorage,Alaska 99501 ABSTRACT Studies of salmon spawning,embryo incubation,and juvenile rearing are all critical in understanding the current life history and habitat dynamics of salmon in the Susitna River.However,the final measure of the value of a reach of river to the freshwater life stages of salmon is the number and condition of the fry which outmigrate from the reach to the ocean.Baseline data on salmon outmigration have been collected at Talkeetna Station (river mile 103.0)for the past three years.The data from 1982 and 1983 have shown that a substantial number of chinook, coho,and sockeye fry outmigrate from the middle river during their first sUlllJler.Because the majority of returning adults have spent at least one winter rearing in freshwater,an important question was whether these age 0+fish overwintered in the lower river or had a low survival rate.To help answer this question,outmigrant traps were also operated near the mouth of the Susitna River (RM 22.4)during 1984. Mark and recapture studies gave population estimates for chum and sockeye fry (marked by coded wire tags)in the Susitna River above Talkeetna Station (middle river)and for chinook fry (marked by cold branding)in Indian River and other rearing sites.The cold branding study also monitored outmigration timing from Indian River and obtained estimates of juvenile chinook residence time in mainstem rearing areas. The Talkeetna River and Deshka River were intermittently sampled to help explain the mainstem outmigrant trap data.A portion of the age 0+ chinook fry apparently outmigrate from the middle river upon reaching a critical size but a large number remain to overwinter and then out- migrate during their second summer.Coho fry outmigrate at a wider range of lengths than chinook fry so the cumulative biomass of coho fry lags behind the cumulative numbers of individuals by one or two weeks. Age 0+chinook and coho fry grow about 30 mm in 1ength duri ng the i -open-water season.Juvenile sockeye salmon appear to seek out lake-like rearing areas at a size of about 50 mm.The limited amount of this habitat type in the middle river is the major influence on their redis- tribution to the lower river.The estimated 1984 middle river population size was about 300 t OOO for age 0+sockeye and 2 t 040,OOO for chum fry.Chum fry rearing in the middle river was demonstrated by their growth and by analysis of stomach contents.~ - - ..... - ,..... ii I." ABSTRACT ••.•••••••••• TABLE OF CONTENTS LIST OF FIGURES ••• ............................................. .'. i vi LIST OF TABLES ••••••••;•••••.••• - ...........'. LIST OF APPENDIX TABLES ••••••••••.•••••••••..•••••• xii xiii 1.0 2.0 I NTROD.UCT ION •••••••••••••••••••••••••••••••••••••••••••'••'•••• METHODS •••••••.••••• 1 3 Study Locations ••••••••2.1 ....................................3-2.1.1 Flathorn Station ••••••...................'.. 2.1.2 Deshka River ...................... 2.1.3 Talkeetna Ri ver .•••.••................ 2.1.4 Talkeetna Station ••••••••............. 2.1.5 Coded wire tagging ••••.............. 2.1.6 Cold branding ........................................ 3 3 3 3 3 10 Recording •••2.2 Field Data Collection and .....................11 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 Flathorn Station outrrtigrant traps ••••••• Deshka River outmigrant weir ••••••.••••. Talkeetna River beach seining •••••••••.••••••• Talkeetna Station outmigrant traps ••. Coded wire tagging •••.•••••••• Cold,branding ••••••••••••••••~•••••••••••••••• 11 11 12 12 12 13 2.3 Data Analysis •••••••.••••••14 3.1 Chinook Salmon . 3.0 RESULTS •••.•••••• 2.3.1 2.3.2 2.3.3 Juvenile salmon catch per unit effort. Population and survival estimates. Time series analysis •••••••••••••••.•••.• ............................................ 14 16 16 17 17 3.1.1 Catch per unit effort...................................17 3.1.L1 3.1.1.2 Age 0+••.•.•••••..••••••• Age 1+•....•••••••.••••.. 17 19 3.L2 Growth •.••••27 3.1.2.1 3.1.2.2 Age 0+•.• Age 1+••. iii 27 27 TABLE OF CONTENTS (Continued) 3.1.3 Cold branding . 3.1.4 Population estimates . Page 27 31 '""" 3.2 Coho Salmon •••••••..•••••..••...•••••••c •••••••••e ••••••••o 35 3.2.1 Catch per unit effort..................................35 3.2 .1.1 Age 0+c ••••••1!I •iii • • • • • • • • •~5 3.2.1.2 Age 1+and older o.........................39 3•2•2 Growt h•••••••••••••••••••••••••••• ••••••••••••••••••••• • 39 3•2•2•1 Ag e .0+D • 3 ..2.2.2 Age 1+and older e •••••••••D. 3.2.3 Cold branding ...•••..........•.•.•.•.•.....••.o ••eoeosc 3.2.4 Population estimates oa •••••••••o.c •••••••• 39 44 44 44 3.3 Sockeye Sa 1mon '00 ••0 •••CI a ••••••••••••0 •eo.••44 3.3.1 Catch per unit effort..................................44 3•3 •1.1 Ag e 0+0 ••••••••••••••••e •••c ••D iii)e e ••e _. 3.3.1.2 Age 1+0 ••••••••c •••c •eo •e III •••••••••• 3 e 3•2 Growt h••••••••••••••••••••••••• •e ••&••II •II • • • • ••D •C ••••• 3.3.3 Coded wire tagging and recovery .•••••••••••••..••..•.•• 3.3.4 Population estimates and survival rates of outmigrants. 47 47 54 54 60 - 3.,4 Chum Salmon _o ••••••••••__._•••••o •••_.60 304el Catch per unit effort ..o•••••_••••ea.oe ••••eo.m •••••••• 3•4•2 Growt h••••CI ••••CI ••• •e ••••••••G CI • • ••••••Ii!•••e eo ••••e _C'••• 3.4.3 Coded wire tagging and recovery ••.•.•••••••••..••••.•.• 3.4.4 Population estimates and survival rates of outmigrants. 60 64 64 64 3.5 Pink Salmon ..__ __..67 3.6 Descriptive Statistics for Catch and Environmental Variables.67 4.0 DISCUSSION...................................................74 4.1 Chinook Salmon __74 _. 4.1.1 Ou tmi gra t ion _C'•eo •••••••••••••e ••••a _••••••0 CI 0 •••• 4.1.2 Freshwater life history •....•••......•..•.•..•.•.•••... 4.1.3 Estimates of population size and residence time .••....• 4.1.4 Growt h__ _ _ _.__ iv 74 74 78 79 - TABLE OF CONTENTS (Continued) 4.2 Coho Salmon................................................79 4.2.1 Outmigrati-on..79 4.2.2 Freshwater life history .••.•.......••••...•..••..~.....83 4 .2.3 Growth.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 - ,~ 4.3 Sockeye Salmon -. 4.3.1 Outmigration · ·. 4.3.2 Freshwater life history .•••...•••.•......•......•...... 4.3.3 Estimates of population size and survival.....•.•....•. 4.3.4 Growth e •••••••III 88 88 88 91 92 4.4 Chum Salmon-••••.•••.••••.•••••••..•••••••••••••-••••••D.....96 4.4.1 Outmigration 10.....96 4.4.2 Freshwater life history................................96 4.4.3 Estimates of population size and survival..............96 4.4.4 Growth.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 4.5 Pi-nk Salmon................................................99 4.5.1 Outmigration --99 4.5.2 Freshwater life history................................99 5.0 CONTRIBUTORS ••••.••_••••••••.••_................................101 6.0 ACKNOWLEDGEMENTS ••••••••.e • • • • • • • • • • • • • • • • • • • • • • • •••• • • • • • • • •103 7.0 LITERATURE CITED •••••••-.......................................104 8.0 APPENDICES Appendix A Juvenile Salmon Catch and Length Data,1984 Appendix B The Schaefer Estimate of Population Size Appendix C Time Series Analysis of Discharge,Turbidity, and Juvenile Salmon Outmigration in the Susitna River~Alaska v L1ST OF FIGURES Figure Title 1 Map of juvenile salmon outmigration study field stations in the Susitna River basin, 1984 "ID •••••ell.If!•.ilI a 4 River,1984 ...............................•0..........9 6 Ma p of coded wi re taggi ng and cold brandi ng sites in the middle reach of the Susitna 2 3 4 5 7 8 9 10 Map of the stationary outmigrant trap and the mobile outmigrant trap sampling points on the Susitna River at FlathornoStatjon,1984 ••.•.•.••••...• Bottom profile of the Susitna River at the stationary and mobile outmigrant trap sampling points at Flathorn Station ••.•.•••....•••.•.. Map showing the location of the fyke net weir on the Deshka River,1984 ••.••...•••.•.•••••.••••••..• Map showing the reach where juveni 1e salmon mark-recapture sites are located (RM 122.2 to 144.8 and Indian River)and the locations of the Talkeetna stationary outmigrant traps (RM 103.0)and the Talkeetna River sampling site (TRM 1.0 ),1984 ....-.. . . . . . •. . . •. •. . . . . . . . . . . ...e ••••S G Branding locations and sample brands used for cold branding chinook and coho salmon j uven i 1es,1984 e •"••••••••IIlI •••••••• Chinook salmon (age 0+)average catch per minnow trap by sampl i ng peri od and survey section in Indian River,1984 •••.•••••.•••.••••••••••• Chinook salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6, 1984.. .ill S •••••"DOD ••••••e • Chinook salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmi grant trap,May 20 through October 1, 1984 0 8 Q •'II •••••••••'"••••10 e ••••• vi 5 6 7 8 15 18 20 21 - - - _. - LIST OF FIGURES (Continued) - - - - Figure 11 12 13 14 15 16 17 18 19 20 Title Chinook salmon (age 0+)daily catch per unit effort recorded at the Fl athorn mobil e outmi grant trap,July 12 through August 30, 1984 . Chinook salmon (age 0+)percent of total catch by sampling point recorded at the Flathorn mobile outmigrant trap,1984 ••••.•.••••.••.•• Chinook salmon (age 0+)catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984 •.••••••••.• Chinook salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Taol keetna stationary outmigrant traps,May 14 through October 6, 1984 ..''.. Chinook salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Fl athorn stationary outmi grant trap,May 20 through October 1, 1984 '' . Chinook salmon (age 0+)mean length and range of lengths by sampling period for fish collected in the lower and middle reach of the Susitn-a River,1984 ·.". Wei9ht/length relationship for juvenile chi nook sa 1mon co 11 ected at the Ta 1keetna stationary outmigrant traps,1984 .•••••••...•••••••... Catch,estimated population size,and main- stem discharge level at Moose Slough,August 8 -August 12,1984 . Catch,estimated population size,and main- stem discharge level at Lower Side Channel llA,July 29 -August 2,1984 .•.•••••••......••.•...•• Coho salmon (age 0+)average catch per minnow trap by sampling period and survey section in Indian River,1984 •••..........•.••.•.•....•..•.•.•... vii 22 23 24 25 26 28 29 33 34 36 LIST OF FIGURES (Continued)'- Figure 21 22 23 24 25 Title Coho salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmi grant traps,May 14 through October 6, 1984 eo C III 0 ..............• '. Coho salmo~(age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Fl athorn stationary outmi grant trap,May 20 through October 1, 1984 0 •••••••••••••••••••••• Coho salmon juvenile catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984 •.•••••••.•..•.• Coho salmon (age 1+and older)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmi grant traps,May 14 through October 6, 1984 II Coho salmon (age 1+and older)smoothed daily catch per unit effdrt and adjusted cumulative catch recorded at the Flathorn stationary outmi grant trap,May 20 through October 1, 1984 ~II-C .. 37 38 40 41 42 - - 26 Coho salmon (age 0+)mean length and range of lengths by sampling period for fish collected in the lower and middle reach of the Susitna R;ve r,1984 ..e •0 0 Gl •G 1&0 e 0 Q Cl g IP "..CD.II •.... ..43 27 28 29 Coho salmon (age 1+)mean length by month for fish collected in the lower and middle reach of the Susitna River,1984 •••••••.•••.••••.•••••.•.••• Weight/length relationship for juvenile coho salmon collected at the Talkeetna stationary outmigrant traps,1984 ............•.................1Il. Sockeye salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6, 1984 .. viii 45 46 48 ~, -. LIST OF FIGURES (Continued) Figure 30 31 Title Page Sockeye salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1, 1984 -0 49 Sockeye salmon (age 0+)daily catch per unit effort recorded at the Flathorn mobile outmi grant trap,July 12 through August 31, 1984 ..,50 32 33 34 Sockeye salmon (age 0+)percent of the total catch by sampling point recorded at the Flathorn mobile outmigrant trap,1984 .•.•••.•....•.••. Sockeye salmon juvenile catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984 .•...•..•... Sockeye salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Fl athorn and Talkeetna stationary outmi grant traps,May 14 through October 6,1984.e_••••••••••'. 51 52 53 35 Sockeye salmon (age 0+)mean length and range of lengths by sampling period for fish collected in the lower and middle reach of the Susitna River~1984 -..55 ...., 36 37 38 39 Weight/length relationship for juvenile sockeye sa lmon co 11 ected at the Ta 1keetna stationary outmigrant traps,1984 •••••.••.•.•••••.••.. Length of time between mark and recapture of coded wire tagged sockeye salmon juveniles in the middle reach of the Susitna River,1984 .••••••.••• Chum salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps, May 14 through October 6,1984 •••.•.•••..•.•••••..•••• Chum salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap, May 20 through October 1,1984 .•••.••.••••••••.•••••.• ix 56 58 61 62 LIST OF FIGURES (Continued) 66 Figure 40 41 Title Page Chum salmon fry catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the.Susitna River,1984................63 Length of time between the mark and recapture of coded wire tagged chum salmon juveniles in the middle reach of the Susitna River,1984 •..•.•.•.•• 42 43 44 45 46 47 48 49 50 Pink salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps, May 14 through October 6,1984 ••••••••••••••••••••••.• Pink salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant traps, May 20 through October 1,1984 ••••••••.••••••••••••••• Mainstem discharge,water temperature,and turbidity in the middle reach of the Susitna River,1984 CG. Mainstem discharge in the lower reach of the Susitna River measured at the USGS gaging station at Susitna Station,1984 •••..••••.•••••••••••. Chinook salmon (age 0+)adjusted cumulative catch recorded at the Talkeetna stati onary outmigrant traps,1983 and 1984 ••••••••••.•••.•••.•••. Chinook salmon (age 1+)adjusted cumulative catch recorded at the Talkeetna stati onary outmigrant traps,1983 and 1984 ••.•••.•.•••••••..•••.. Chinook salmon (age 0+)mean length and range of mean 1engths by sampl ing period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984 ••••••••.•.••••••••...•.••• Chinook salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and Flathorn stations,1984 ••••.••••.••.•..•...•••.•.• Coho salmon (age 0+)adjusted cumulative catch recorded at the Tal keetna stationary outmigrant traps,1983 and 1984 ••........•.•••••.•.••• x 68 69 72 73 75 76 80 81 82 - - LIST OF FIGURES (Continued) Figure Title - - - 51 52 53 54 55 56 57 58 59 60 61 Coho salmon (age 1+)adjusted cumulative catch recorded at the Talkeetna stati onary outmigrant traps,1983 and 1984 ••....•••.•...•••.••••. Coho salmon (age 0+)mean length and range of mean 1engths by sampl i ng period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984 ••••••••.••••...••••••.•••• Coho salmon (age 1+)mean length and rang·e of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982, 1983,and 1984 .•..••••.•••••••••.•••••••• Coho salmon adjusted cumul ative catch and bi amass by age class recorded at Talkeetna and Fl athorn Stati ons,1984 ••••••••••••••••••.•.•••.•. Sockeye salmon (age 0+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984 .•••••••••••••.•••••.•• Mean length of coded wire tagged sockeye salmon fry at recovery sites in,the middle reach of the Susitna River by week,1984 .••••.•••••.•• Sockeye salmon (age 0+)mean length and range of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984 •••••••••••••.•.•••.••••.•• Sockeye salmon adjusted cumulative catch and biomass by age cl ass recorded at Talkeetna and Flathorn Stations,1984 •..•••••••••.••••••.•.•.••• Chum salmon fry adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984 ••••.••..•.••••..••..•. Mean length of coded wire tagged chum salmon fry at recovery sites in the middle reach of the Susitna River by 5 day period,1984 •.••••••.•.••.• Pink salmon fry adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984 ••.••...•....•.••••.... xi 84 85 86 87 89 93 94 95 97 98 100 LIST OF TABLES - Table 1 Title Page The number of chinook salmon fry marked and recovered in Indian River by sampling period, 1984 ~e ••••••••IZI •e •00 3a - 2 3 4 5 6 7 8 Chinook salmon fry population estimates by site for middle Susitna River sloughs and side channels and for Indian River,1984 •..•.•••••••.• Coded wire tag release data for sockeye salmon fry on the Susitna River by tagging site and release date,1984 .•••••••••.•.••••••••••.••. Recoveries of coded wire tagged sockeye salmon fry at mainstem river sites between Talkeetna and Devil Canyon,1984 •••••••••••••••••••••• Coded wi re tag release data for chum salmon fry on the Susitna River by tagging site and re 1ea se da te·".1984 •••••••••••••••••Go •Co •••••••••••••••III Summary statistics for juvenile salmon catch per hour by species and age class recorded at the Talkeetna Station outmigrant traps,May 14 through October 6,1984 •••••••••••••••••.••••••••.• Summary statistics for habitat variables recorded on the Susitna River between the Chul i tna Ri ver confl uence and Devi 1 Canyon, May 14 through October 6,1984 ••••••••.••••••••••••••• Summary statistics for juvenile salmon catch per hour by species and age class recorded at the Flathorn Station outmigrant traps,May 20 through October 1,1984 . xi i 32 57 59 65 70 70 71 - ~I LIST OF APPENDIX TABLES Appendix Table Title A-1 Wei r catches of juveni 1e chi nook and coho salmon on the Deshka River,May 10 through September 19,1984....................................A-1 A-2 Results of incidental minnow trapping in the Deshka River,1984....................................A-2 A-3 Mean 1ength and range of lengths for age 0+ chinook salmon by sampling period in the lower reach of the Susitna River,1984................A-3 A-4 Mean length and range of lengths for age 0+ chinook salmon by sampling period in the Tal keetna River and the middle reach of the Susitna River,1984 ...~...............................A-4 Mean length and range of lengths for age 0+ coho salmon by sampling period in the lower reach of the Susitna River,1984......................A-6 Mean length and range of lengths for age 1+ chinook salmon in the Susitna River,1984.............A-5 Mean length and range of lengths,for age 0+ coho salmon by sampling period in the middle reach of the Susitna River,1984......................A-7 A-5 .- A-6 r- A-7 A-8 Mean 1ength and range of 1engths for age 1+ coho salmon by sampling period in the lower reach of the Susitna River,1984......................A-8 - .... ...... A-9 A-I0 Mean length and range of lengths for age 1+ coho salmon by sampling period in the middle reach of the Susitna River,1984......................A-9 Mean length and range of lengths for age 2+ coho salmon by sampling period in the Susitna Ri ver between Cook In 1et and Dev i 1 Ca nyon , 1984 •.•.•...•••...••...•••••.•.••••...•.••..•..•••.••.A-I0 A-ll Daily catches of outmigrant chum and sockeye salmon fry in a fyke net located at the mouth of Slough 21,May 23 to June 12,1984.................A-ll A-12 Mean 1ength and range of 1engths for age 0+ sockeye salmon by sampling period in the Susitna River between Cook Inl et and Devi 1 Canyon,1984...........................................A-12 xiii LIST OF APPENDIX TABLES (Continued) Appendix Table Title A-13 A-14 Mean 1ength and range of 1engths for age 1+ sockeye salmon by sampling period in the Susitna River between Cook Inl et and Devi 1 Canyon,1984 -o ••••••••o8 0Iloo ••e.-o.A-13 Mean 1ength and range of 1engths for chum salmon fry by sampling period in the Susitna River between Cook Inlet and Devil Canyon, 1984.........•••••• ••.•.•••••••.•..•.•••.•.•••••••.•••A-14 ..... B-1 Data collected on the coded wire tag,mark- recapture experiment for sockeye salmon fry to prOVide a Schaefer population estimate.............8-2 B-2 Computati on of the sockeye sa 1man for outmigrant population from the data presented in Appendix Table B-l ••.•••••e •••••••~•••••e.e •••ee.8e B-3 B-3 Data call ected on the coded wi re tag,mark- recapture experiment for chum salmon fry to provide a Schaefer population estimate................8-4 8-4 Computation of the chum salmon for outmigrant population from the data presented in Appendix Table B-3 ••••••••~•.•••e.~~••••e$••o~o ••••o..B-5 xiv - ~I - """' 1.0 INTRODUCTION Studies of the migration and growth of juvenile salmon in the mainstem Susitna River are a part of the ongoing investigations being conducted by the Resident and Juvenile Anadromous Fish Project (RJ)of the Susitna River Aquatic Studies Program.The scope of these studies has been to describe the periods of freshwater residence,growth,and timing of outmigration for juvenile salmon in the Susitna River and to provide population estimates for the reach of river between the Chulitna Riv~r confluence and Devil Canyon.This report presents the results of juvenile salmon outmigration stl,ldies conducted on the Susitna River between Cook Inlet and Devil Canyon during the 1984 open-water season. Five Pacific salmon species are addressed in this report:chinook (Oncorhynchus tshawytscha),coho (Q.kisutch),sockeye (0.nerka),chum (Q.keta),and pink (0.gorbuscha). Investigations of the distribution,abundance,and migration of juvenile salmon during 1982 and 1983 were focused primarily on the Susitna River reach above the Chulitna River confluence (ADF&G 1983a;Schmidt et ale 1984).These studies included the operation of stationary outmigrant traps at Talkeetna Station,river mile (RM)103.0,during 1982 and o1983 and a mark-recapture program for post-emergent chum and sockeye sal mon fry using half-length coded wire tags in 1983 (Roth et ale 1984).These techniques have provided valuable information on the success of previous spawning runs,the effect of di scharge on redi stri bution of young-of- the-year salmon juveniles,and the population size and egg-to-outmigrant fry survival rates for chum and sockeye salmon fry. During the 1984 open-water season,additional tasks were added to further describe juvenile salmon growth,migration timing,and response to changing habitat conditions.The study area was expanded to include the entire river between Cook Inlet and Devil Canyon.New tasks begun in 1984 were the addition of stationary and mobile outmigrant traps at Flathorn Station (RM 22.4),intermittent trapping of migrating chinook sa 1mon juven i1es in the Deshka and Talkeetna ri vers,and mark-recapture by cold branding of juvenile chinook and coho salmon in the Curry Station to Devil Canyon reach. Investigations of the migration and growth of juvenile salmon in the Susitna River above the oChul itna River confl uence duri ng 1982 and 1983 indicated extensive migration of pre-smolt juveniles of all species to areas below this reach.This migration of pre-smolt chinook salmon was also observed in the Deshka River in 1980 (Delaney et ale 1981).If this movement is common in the major tributaries entering the Susitna River, extensive rearing and growth of juveni1e salmon,parti cul arly chinook, may occur in habitats associated with the mainstem river.Small habitat changes in the reach of river below Talkeetna could impact large numbers of rearing salmon. The combined studies of juvenile salmon growth and migration conducted during the 1984 open-water season were developed to provide data to meet the following objectives: 1 o Estimate the timing,relative abundance,and size of out- migrating juvenile salmon in the Susitna River above the Chulitna River confluence. o Estimate the population size of outmigrating chum and sockeye salmon fry and egg-to-outmigrant fry survival in this reach of .river. o Estimate the timing and size of outmigrating chum salmon from the Talkeetna River. o Estimate the timing and rate of movement of juvenile chinook and coho salmon out of Indian River and their residence time at selected macrohabitats associated with the mainstem Susitna River. - o o o Estimate the timing and rate of outmigration of chinook salmon juven'iles from the Deshka River into/the mainstem Susitna. ! Estimate the timing and rate·of \outmigration of juvenile salmon from the Susitna River into Coo~Inlet. Estimate the rate of growth of juvenile chum and chinook salmon from the time they enter the lower river (below the Chulitna River confluence)until they enter the marine environment. o Estimate the relationship of mainstem Susitna discharge and other environmental variables to juvenile salmon outmigration. Sampling of chum salmon fry in the Talkeetna River was hindered by equipment failure;insufficient data were collected for this species, although some growth and relative abundance data for chinook salmon were collected. Although initially designed as a survey of Portage Creek using a sta- tionary outmigrant trap,the cold branding study was relocated to Indian River with minnow traps serving as the primary collection technique. The design of the original collection equipment did not lend itself well to the continually fl uctuating hydraul ic conditions present at Portage Creek.The low numbers of juvenile salmon observed in Portage Creek after June 15,combined with the comparative logtstical inaccessibility of this stream,made Indian River a better choice. The data presented in this report provide information that can be used to determine the size of the present fishery resource,potential changes caused by the proposed hydroelectric development,and mitigation requirements necessary to compensate for any reductions of the juvenile salmon populations in the Susitna River. 2 .- 2.0 METHODS 2.1 Study Locations Studies on the migration and growth of juvenile salmon in the mainstem Susitna River were conducted at survey sites from Flathorn Station (RM 22.4)upstream to Slough 22 (RM 144.3)during the 1984 open-water season (Fig.1). 2.1.1 Flathorn Station A stationary outmi grant trap was operated on the west bank of the Susitna River at Flathorn Station (RM 22.4)and a mobile outmigrant trap was used to sample a total of ten points along transects spanning three channels of the mainstem river at this station (Fig.2).Five sampling points were located in the west channel (RM 22.8),one in the middle channel (RM 22.8),and four in the east channel (RM 23.9).A bottom profile of the Susitna River at these sampling points is provided in Fig.3. 2.1.2 Deshka River An outmigrant fyke net weir was operated in the Deshka River (RM 40.6) between tributary river mile (TRM)2.5 and TRM 5.0 to estimate the timing and rate of outmigration for juvenile chinook salmon (Fig.4). 2.1.3 Talkeetna River A beach seine sampl ing site for outmigrants was located in the north channel of the Talkeetna River (RM 97.5)approximately one mile upstream from the river's mouth (Fig.5). 2.1.4 Talkeetna Station Two stationary outmigrant traps were deployed on the mainstem Susitna River above the Chulitna River confluence at Talkeetna Station (RM 103.0)at the same locations used in 1983.One trap was set off the east bank (Trap 1)and the other off the west bank (Trap 2)of the river (Fig.5). 2.1.5 Coded wire tagging Coded wire tagging sites were selected from those locations above the Chul itna River confluence where hi gh density spawning by adults was recorded (Barrett et ale 1984),and from surveys of the availability of sufficient numbers of post-emergent chum and sockeye salmon fry for collection and tagging (Fig.5).Specific coded wire tagging sites (Fig. 6)were: 3 TALKEETNA~~~la STATION ~ 0 10 20 30 I I r I MILES (Appro •.Scale) / ~; - - Figure 1.Map of juvenile salmon outmigration study field stations in the Susitna River basin,1984. 4 - SAMPLING POINTS •Stationary Outmigrant Trap •Mobile Outmigrant Trap 4 ,'.;. ~... ,f I">• " ~~ '".?:V, ~ CI) ...... "f ~ ~".~ ~." ".' '=...!~ ~ ~ J " .' ~. FlATHORN STATION .- - Figure 2.Map of the stationary outmigrant trap and the mObile outmigrant trap sampling points on the Susitna River at Flathorn Station t 1984. 5 EAST CHANNEL IHM23.91 ~IICW 0 ••••0 e ••-iE Go '0l!f l' 1.l'I 11.110 '"~100 .1'710 .71 'GOO 111. HORIZONTAL DISTANCE ern ,~ - IIDDLE CHANNEL IBI 2%.81 o • SAM PLINK POINTS x STATIONARY OUTMlGAANT TRAP •MOBLE OUTMIGRANT TRAP 10 10 o 111 110 17. HORIZONTAl.DISTANCE am WEST CHANNEL IBM 22.41· - o • Raw••• • • f l'-% l- Go ~,. I.+---r-......,r---~-....,..---r--~-....,..---r--.,..---r----r--+-tl o 11.110 .7.too II'710 11.'000 "1'',10 '11'1100 HOAIZONTAL DI8TANeE (FT) Figure 3.Bottom profile of the Susitna River at the stationary and mobile outmigrant trap sampling points at Flathorn Station.Measured on August 23,1984 at a mainstem discharge of 114,000 cfs at the· USGS gaging station at Susitna Station. 6 MILES (ApprOll.Scale)fJ eRM45 Fyke Net------~n Weir Site (TRM 2.5) r- ! - Figure 4.Map showing the location of the fyke net weir on the Deshka River l 1984. 7 - ~ 0 10 I I I MILES ~Coded Wire Tagging a Cold Branding Sites "'0 .....0'"co ,.••,f "-.-Talkeetna River Sampling Site Talkeetna Station Out migrant Traps Figure 5.,Map showing the reach where juvenile salmon mark-recapture sites are located (RM 122.2 to 144.8 and Indian River)and the locations of the Talkeetna stationary outmigrant traps (RM 103.0),and the Talkeetna _ River sampling site (TRM 1.0),1984. 8 •SAMPLING SITE LOUGH 21 SIDE CHANNEL SLOUGH 21 SLOUGH 20 SLOUGH 19 SLOUGH II UPPER SIDE CHANNEL II SI DE CHANNEL 10_ SLOUGH 9 SLOUGH 8A MOOSE SLOUGH SLOUGH 88 R. SLOUGH 22 INDIAN RIVER SLOUGH 17 SLOUGH 16 SLOUGH 15 4t11 of JulyCr. __'"'==I-TALKEETNA STATION OUTMIGRANT TRAPS ..... Figure 6.Map of coded wire tagging and cold branding sites in the midole reach of the Susitna Rivers 1984. - 9 CODED WIRE TAGGING SITES RIVER MILE Slough 8B 122.4 Slough 8A 125.3 Slough 9 129.2 Slough 11 135.3 Slough 15 137.3 Indian River 138.6 Slough 20 140.1 Slough 21 142.0 Slough 22 144.3 2.1.6 Cold branding A col d brand mark-recapture study was conducted at the mouth and at numerous side channels and side sloughs of Indian River (RM 138.6)which were found to contain large concentrations of juvenile chinook and coho salmon.Indian River was divided into three sections for this study. Section I included the mouth upstream to TRM 0.5,Section II was the portion of Indian River from TRM 0.5.to 7.5 and Section III was from TRM 7.5 upstream to TRM 12.3 (Fig.5). Cold branding was also used to estimate the populations and study the movements of juvenile salmon at the following study sites (Fig.6): COLD BRANDING SITES Moose Slough Side Channel 10 Lower Side Channel llA Slough 16 Slough 17 Slough 19 Slough 20 Side Channel 21 Slough 22 10 RIVER MILE 123.2 133.8 135.9 137.7 138.9 139.7 140.1 141.1 144.3 - -- 2.2 Field Data Collection and Recording 2.2.1 Flathorn Station outmigrant traps The stationary outmigrant trap on the west bank of the Susitna River at Flathorn Station (RM 22.4)was operated from May 20 through October 1, 1984.A description of this outmigrant trap is provided in ADF&G (1985).The trap was checked at least twice each day to remove the captured fish and to clean the trap. The mobile outmigrant trap.at Flathorn Station was.operated for 43 days during the period July 12 through September 13,1984.A description of the trap and its operation is presented in ADF&G (1985).The trap was fished for 20-minute periods at ten different transect points during a fishing day. Habitat and biological data recorded for each check of the stationary outmigrant trap included fishing effort (hours),trap depth (feet), distance from shore (feet),and catch by species and age class.Main- stem stage was recorded once each day.The fi rst 25 fi sh of each species and age class collected daily were measured for total length (tip of snout to tip of tail)in millimeters (mm). Biological and habitat data for the stationary trap were entered directly into an Epson HX-20 microcomputer in the field.·Operational procedures f.o..r the microcomputer and the associated data form program are presented in ADF&G (1985).Computer entri es were made for each trap check throughout the fi~ld~eason.Printouts and cassettes were periodically transferred to Data Processing to be entered into a main- frame computer for later data retrieval and analysis. Transect number,fishing effort,total water column depth,set velocity, and drift velocity (if the trap was not held stationary during the set) were recorded for each individual transect point at which the mobile outmigrant trap was fished.Total catch by species and age class was also recorded,and total length measurements were taken for all captured fish.Data were recorded on a field data form for later analysis. 2.2.2 Deshka River outmigrant weir A weir was established on the Deshka River (RM 40.6)using a fyke net (3/16 inch square mesh)to block a portion of the river.The fyke net is described in ADF&G (1985).The weir was operated at varying tributary miles (TRM 2.0 -5.0)periodically from May 10 through June 22.The weir was moved to TRM 2.5 on July 11 and was fished periodically through September 18.Minnow traps were fished intermittently from 1ate June through mid October to supplement the weir data. Fishitlg effort and total catch by species and age class were recorded for the outmigrant weir and the minnow traps.A sample of each species and age class captured were measured for total length and scale samples were collected for age determination. 11 2.2.3 Talkeetna River beach seining Beach seining (1/8 inch square mesh)was conducted one to two times each week from June 5 through September 15.Sampling was conducted to obtain a sufficient sample for comparative length and outmigration timing data. An attempt was made to use a Fyke net weir in late May and June.This did not work,so we changed to a beach seine. Total catch by species and age class was recorded.All.captured fish were measured for total length and released. 2.2.4 Talkeetna Station outmigrant traps Two inclined plane outmigrant traps were operated continuously in the mainstem Susitna River at Talkeetna Station (RM 103.0)from May 14 through October 6,1984 using the methods outl ined by Roth et al. (1984). Measurements of the following habitat parameters were recorded daily at the outmigrant traps:air and surface water temperature (OC),turbidity (NTU),water velocity (ft/sec),and mainstem stage data.The equipment and methods used to collect the habitat data are given in ADF&G (1985). Trap fishing depths and distances from shore were adjusted to maximize catches while maintaining trap efficiency.All juvenile fish captured were anesthetized using MS-222 (Tricaine methanesulfonate).Field specimens were identified using the guidel ines set forth by McConnell and Snyder (1972),Trautman (1973),and Morrow (1980).Juvenile chinook and coho salmon collected at the traps were checked for a cold brand mark and all recovered marks were recorded.Chum and sockeye salmon juveniles with a clipped adipose fin were passed through a detector to veri fy the presence of a coded wi re tag.A11 coded wi re tagged fi sh recovered at the traps were preserved and tags were 1ater removed and decoded using a reading jig and a binocular microscope.All other fish recovered at the traps were held until anesthetic recovery was complete and then released downstream of the traps. Scales were collected from a sub-sample of fish captured for comparison to length frequency data for final age class determination.Biological and habitat data were entered directly into an Epson HX-20 mi crocomputer. Length and weight relationship data were also collected from samples of juvenile chinook,coho,and sockeye salmon collected in the outmigrant traps at Talkeetna Station.Total length was recorded to the nearest millimeter and live weights were determined to the nearest 0.1 gram. 2.2.5 Coded wire tagging The coded wire tagging was conducted at Slough 11 (RM 135.3)from May 16 through June 20,1984.The fish were transported from the collection areas to Slough 11 in an aerated tub,tagged,held for at least 24 hours,and then returned to the collection areas •.The fish were also held overnight at the collection areas prior to release. 12 ...., - - - -c ~ I - Beach seines were used to weir off the downstream end of the collection area and were checked at least once each day to collect fish and remove debris.Beach seining and dip netting supplemented the weir catches at sites where wei ring alone did not provide enough fish for the tagging operation. The coded wire tagging equipment and implantation procedures are similar to those outlined by Roth et a1.(1984)using the guidelines provided by Koerner (1977)and Moberly et al.(1977).One,;"ha1f length binary coded wire tags measuring 0.02 inches (0.533 mm)in length and 0.01 inches (0.254 mm)in diameter were used in the study.Separate head molds were required for each species and length class of fish.Fifty fish of each group were measured to determine mean length and the proper head molds for the tagging procedure.The adipose fin was clipped from each fish prior to tagging to provide a visual indicator of the presence of a coded wire tag.At the end of each tagging day,a subsamp1e of 100 tagged fish were anesthetized and passed through the quality control device to determine the tag retention rate.Mortalities were recorded the following day and again just prior to release.A single tag code was used for each species tagged and for each collection site.Six distinct tag codes were used for juvenile sockeye salmon and fourteen distinct tag codes were used for juvenile chum salmon. Coded wire tagging data recorded at each site included date tagged,tag code,speci es,number of fi sh tagged,percent tag retent i on,mortal i ty, and date and time of release.Total numbers of fish tagged by species, collection site,and release date as well as final tag retention and mortality were tabulated for each tag code. 2.2.6 Cold branding Mark-recapture studies of chinook and coho salmon populations were conducted from July through mid October.Cold branding was used as a marking technique because it is less expensive than coded wire tagging. Cold branding was not used on chum and sockeye because it has not been proven effective on these fish at the post-emergent stage.Sites in Indian River were sampled twice a month and fish were captured,branded, and released continually throughout the field season.Sampling in the sloughs and side channel s of the Susitna River was conducted for five consecutive days and captured fish were either branded and released the same day or held until the end of the five day period before release. Minnow traps,beach seines,and dip nets were used to capture fish which were then transported from the areas of collection to the Gold Creek field camp for cold brand marking.Cold branded fish from all sites except Indian River were held for 24 hours to determine marking mortality before being released at the area of collection.Fish col- lected in Indian River were marked,held for 24 hours,and then released at a side slough at TRM 7.2. The brands consisted of single brass letters or symbols measuring approximately three millimeters in height which were soldered onto threaded brass caps.Liquid nitrogen was used as the cooling agent and 13 branding procedures were similar to those outlined by Raleigh et al. (1973).The cold branding equipment is described in ADF&G (1985). Juvenile chinook and coho salmon were marked with a distinctive brand to signify the collection site and date of their capture.Fish were marked on one side of th~body at one of three target branding areas (Fig.7), and a branding time of two seconds was used.- Date,collection site,gear type,fishing effort,species,number of fish captured,and brand symbol were recorded for each site.The number .~ of recaptures by species and the symbols for previously marked fish were also recorded.Total length was measured for 50 fish of each species during each sampling trip. 2.3 Data Analysis 2.3.1 Juvenile salmon catch per unit effort The catch per unit effort (CPUE)data collected for juvenile salmon at the stationary outmigrant traps are presented as the average catch per hour for each calendar day of sampling effort.The catch was expanded to 24-hour intervals by dividing the number of hours fished on a given day into 24 and then multiplying this ratio by the catch for each species and age class. The catch rates plotted for each species and age class of juvenile salmon collected at the stationary traps were smoothed using the von Hann linear filter (Dixon et al.1981).The equation is: Z(t)=iY(t_1)+iY(t)+iY(t+1).~ where:Z(t)=smoothed catch per hour for day (t)and Y(t)=observed catch per hour for day (t) This is similar to a three day moving average except that the current day is weighted twice as heavily as the preceding and subsequent days. The cumulative catch totals were adjusted for days not fished by tabu- lating the mean of the total catches recorded for the three days preceding and the three days following an unsampled day. Length frequency distribution and scale analysis data were used to determine the age class composition of chinook,coho,and sockeye salmon juveniles. A regression was done on the natural logarithm of weight versus the natural logarithm of length for chinook,coho,and sockeye salmon.The regression equations were used to provide estimates of the total biomass passing the Talkeetna and Flathorn station outmigrant traps by sampling period through the season. 14 Six·8 randing Locations Left Side Right Side a)anterior to dorsal fin b)beneath dorsal fin c)posterior to dorsal fin ,..,., I Sample Cold-Brands U ::::>n c ,..., 3EI.LI ITt T I-.L ~ A L 1....J r I -L--1 S -Figure 7.,_Branding locations and sample brands used for cold branding chinook and coho salmon juveniles,1984. 15 2.3.2 Population and survival estimates Potential egg deposition for chum and sockeye salmon was calculated by multiplying the average fecundity for each species by the estimated number of female spawners that passed Curry Station in 1983 (Barrett et al.1984).The chum,sockeye,and chinook salmon adult population estimates were reduced by 40%,39%,and 7%respectively,to account for milling fish which eventually spawned below the Chulitna River confluence (Barrett 1984;Barrett et al.1984).The following formula was used to determine egg deposition: Total potential egg deposition =(E)x (l-M)x (p)x (F) where: E =Adult population estimate at Curry Station M=Proportion milling P =Proportion females F =Average fecundity Population estimates for chum and sockeye outmigrants were calculated by the Schaefer (1951)method (Appendix B).Estimates of survival for both species were determined by dividing the population estimates by the calculated potential egg deposition for each species.Only valid tagged fish were used in the calculations.The total number of valid tagged fish was determined by subtracting the mortalities for each day of tagging from the total number of fish tagged and then multiplying this by the tag retention rate.Total tag recoveries at the Talkeetna Station outmigrant traps include only those fish with a coded wire tag. Fish having a clipped adipose fin but no tag were not considered in the population estimates. Population estimates for chinook salmon were calculated from the data collected during the cold branding study by using the Petersen, Schaefer,or Jolly-Seber methods (Ricker 1975).The Schaefer and Jolly-Seber methods were used at sites where conditions allowed five consecutive days of sampling.The Peterson method was used when there was one marking period and one recapture period.Confidence limits for the Jolly-Seber estimate of population size were developed using the method of Manly (1984).The Jolly-Seber model was run on a commercial spreadsheet program for microcomputers.The potential egg deposition for chinook salmon in Indian River was determined-using the technique 1i sted above except that the estimate was reduced to represent the percentage of chinook (determined from peak spawning counts)which spawned in Indian River.Fecundities used were those measured by Healy and Heard (1984)for Kenai River chinook salmon. 2.3.3 Time series analysis The 1983 and 1984 discharge,turbidity,and age 0+chinook and sockeye salmon outmigration time series are analyzed in Appendix C. 16 - ..... -. - ,.... """, - .- ..... 3.0 RESULTS The results of the juvenile salmon outmigration studies are presented by species.The catch per unit effort (CPUE)data are presented as a percentage of the highest CPUE (after smoothing)recorded at the sta- tionary traps during 1984.The cumulative catch data are presented as a percentage of the total adjusted ·cumu1ative catch after app1 ication of the smoothing functions.Juvenile salmon length data collected at Flathorn Station are from both the stationary and mobile traps and the length information presented for Talkeetna Station is from both stationary traps located at this site. 3.1 Chinook Salmon 3.1.1 Catch per unit effort 3.1. 1.1 Age 0+ Chinook salmon fry collected incidentally during the coded wire tagging study in May and June were observed to be most abundant at Slough 22 and Indian River. The cold branding study captured 26,823 chinook salmon fry in Indian River from July 1 through October 15.Fifty-eight percent of this catch was recorded near the mouth of the river (section I),30%in the lower portion (section II)and 12%in the upper portion (section III).Beach seining of sections II and III during July captured 3,280 chinook salmon fry;66%in section III and 34%in section II.Minnow trapping begun in Indian River in late July collected a total of 23,543 chinook fry during 947 minnow trap days (defined as one trap day for each overnight minnow trap set)for a season average of 24.9 fish per trap day. Catch rates in Indian River (Fig.8)were generally highest in section II except during late August when high and turbid water conditions reduced trapping effectiveness.The CPUE for chinook fry in Indian River for all sections combined was highest during late July (average of 36 fish per trap day)and steadily declined through the season to a low of 15 fish per trap day in early October . A total of 11,875 chinook salmon fry were captured in sloughs and side channels in the middle reach of the Susitna River during the cold branding ~tudy from July 1 through October 15.Sloughs accounted for 84%of the catch while the remaining 16%were collected in side channels.Beach seining during July and August collected 39%of the total catch at these sites while minnow trapping begun in early September captured 61%of the chinook fry. The 7,291 chinook salmon fry captured by minnow trapping at slough and side channel sites in the middle river were collected during 378 minnow trap days for an average of 19 fish per trap day.Mean CPUE by study site ranged from a hi gh of 48 fi sh per trap day at Slough 22 during early October to a low of 3 fish per trap day at Side Channel 21 in late September. 17 INDIA~l RIVER 1 984 - - E OCTLSEPESEP SAlVIPLlNG PERIOD CHI~JOOK 0+ 45 40 35 ~a a.30 «a:: f- a::25 wa. :I:20(j ~ (j '15 10 5 L JUL E AUG L AUG Figure 8.Chinook salmon (age 0+)average catch per minnow trap by sampling period and survey section in Indian River,1984. 18 - - A total of 14,110 chinook salmon fry were collected at the Talkeetna Station outmigrant traps.Peak catches were recorded from late June through early August and the highest catch rate of 17.3 chinook fry per hour was recorded on July 26 (Fig.9).Fifty percent of the catch was recorded by July 20.Catches decreased after early August and the last capture of chinook fry at this site was recorded on September 29. A total of 2,118 ch"jnook salmon fry were captured in the stationary outmigrant trap at Flathorn Station.Catch rates were greatest between late June and late August (Fig.10).The chinook fry catch rate at this site peaked at 7.8 fish per hour on July 23,50%of the captures were recorded by July 13,and the last capture was recorded on September 30. The highest catch rate of the Flathorn Station mobile trap was 16.2 fish per hour,recorded on July 23 (Fig.11).Of the 189 chinook fry coll ected in the mobile trap duri ng 1984,60%were captured at bank transect sampling points and the remaining captures occurred at center channel sampling sites (Fig.12). The Deshka River weir captured 1,808 chinook salmon during 1984 (Appen- dix Table A-I).Eighty-eight percent of the captures were recorded during July and the peak catch rate of 21.2 "fi sh per hour was recorded on July 25.Minnow trap catches at this site were highes"t during late June at 8.7 fish per trap (Appendix Table A-2). A total of 1,356 chinook salmon fry were collected in the lower reach of the Susitna River by the Juvenile Aquatic Habitat Studies (JAHS)surveys from June through early October (see Part 2 of this report).Catch rates for all sites combined peaked in August and then decreased through early October (Fig.13). 3.1.1.2 Age 1+ Age 1+chinook salmon were captured incidentally during the coded wire tagging study in May and June and were most abundant at Indian River and Slough 11.No age 1+chinook were captured during the cold branding study begun in July,as most of these fish had outmigrated by that time. Peak catch rates of the 1,321 age 1+chinook captured at the Talkeetna Station outmi grant traps we·re recorded during the deployment of the traps in mid May and again in mid and late June (Fig.14).Fifty percent of the season catches occurred by June 23.The highest catch rate for this age class was 3.6 fish per hour recorded on May 15 and the last age 1+chinook was captured in the traps on August 7. Catch rates for the 346 age 1+chi nook salmon captured at Fl athorn Station were highest during early June (Fig.15).The highest CPUE of 6.4 fish per hour was recorded on June 14 (50%of the season total by this date)and the last age 1+chinook was collected at this site on August 23. Nine age 1+chinook salmon were collected in the Deshka River during weir and minnow trap sampling,with the last capture recorded on October 10. 19 TALKEETNA BOTH TRAPS ~ Z l.U ()a::w Q. 100 ~ 90 80 170 60 50 40 30 20 10 o~IMAY \ I %CUMULATIVE %OF HIGHEST CPUE ~ AUG r SEP ~14.8 [ It ~ t@ L ~ I..J..O.O Figure 9.Chinook salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. 20 FLATH OR !'J STATION CHINOOK 0+ a:: ;:) o :::I: ....... :::I: U r-« u -%CUMU LATIVE -%OF HIGHEST CPUE ...,------·--------r-------~~----__r_5.1 A-+-----f--"':::::-...,.---+------+--------:::~--::=...-----""=t_o.o JUL SEP' 100 gO 80 70 50I- Z W (J 50ccwa. 40 30 20 10 0 MAY Figure 10.Chinook salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. 21 - AGE 0+CH I hj()(]~<CF'lJE -, 100 -a'-Notsa~~~:--l90 ~b -No Fish Captured w 80 ::} 11. ()70 I- tl) W I 60 I~.,1\;,. 1: lJ..50 0 W G 4-0 ,-~ Z lJJ 30u I:r \.LJ .-ll.20 10 -i b~~.~~1~~" 0 aa aa 12 15 20 JULY 30 5 10 AUGUST 20 --'C.":;·0L,l DATE Figure 11.Chinook salmon (age 0+)daily catch per unit effort recorded at the Flathorn mobile outmigrant trap,July 12 through August 30,1984. 22 AG"E 0+CHI!'JOOK SEASON CATCH-21 20 19 "18 17 :I:16 ()15I-«14(.) ...J 13 ~12 0 11I- 1.1...10 0 9I- Z 8w 0 7 0::6wc..5 4 3 2 1 0 "'1 2 3 4 "'5 "'6 "'7 8 9 "'10 Figure 12. TRANSECT POINT NUMBER *BANK TRANSECTS Chinook salmon (age 0+)percent of total catch by sampling point recorded at the Flathorn mobile outmigrant trap,1984. 23 CHINOOK'CPLIE 1"984 100 -,.....-----------...,..,.......,......,....,--------------, 90 W :::::l 0-o t;j W :I: C) :I: u..a ~ 80 70 60 50 40 r-,,......,......,., 30 20 10 E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLI NG PERIOD - - Figure 13.Chinook salmon (age 0+)catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River, 1984. 24 - - TALKEETNA 80TH TRAPS CHII\-IOOK ~l + %CUMU LAT1\1E %OF HIGHEST CPUE v\_---+_------11.o JUL -4 AUG SEP IJUN 20 60 50 30 70 40 80 90 1:X _ ,MAY l- Z LlJo 0:: W 0.. Figure 14.Chinook salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. 25 FLATHORN STATION CHINOOK 1 + T----,-----==:::::::;:=====-----------r3 .3 - - a:: ~ o ::t:..... ::t: U... 0< u SEPAUG %CUMULAnVE %OF HIGHEST CPUE +........L..1f-------,-___t'-..:.....:=--=~....:=...-o<:::>t--==---=--___t'-------+-O.O 100 90 80 70 60 I- Z W 0 50~wa. 40 30 20 10 0 Figure 15.Chinook salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. - 26 ...., - - ,- -- 3.1.2 Growth 3.1.2.1 Age 0+ Chinook fry collected between the Chulitna River confluence and Devil Canyon (middle river)averaged 43 mm during late May and showed a steady growth through the season to a mean 1ength of 64 mm by early October (Fig.16).Age 0+chinook collected between Cook Inlet and The Chulitna River confluence (lower river)during the same period averaged consistently larger than fry collected in the middle river.Chinook fry in the lower river increased from a mean length of 41 mm in late May to 75mm in early October.The number of fish measured,mean length,and range of lengths by sampling period for chinook salmon fry are presented for each data collection area in Appendix Table A-3 and A-4. 3.1.2.2 Age 1+ Age 1+chinook salmon for all sites sampled averaged 78 mm during May and the mean length increased to 90 mm during early June (Appendix Table A-5).Average 1engths for thi sage cl ass stayed the same through 1ate July by which time most of the age 1+chinook had migrated out of the Susitna River. The length/weight relationship of juvenile chinook (both age classes)at Talkeetna Station is shown in Fig.17. 3.1.3 Cold branding A total of 23,406 chinook salmon fry were cold branded in Indian River between July 1 and October 15,1984 (Table 1).One hundred forty-seven of these marked fish were later recaptured in Indian River,five were captured in the Talkeetna Station outmigrant traps,and five were captured below Indian River in side channels and sloughs associated with the mainstem Susitna River.The time between release of marked chinook fry in Indian River at TRM 7.2 and their subsequent recapture at the mouth of this tributary ranged from nine to 70 days with a mean of 30 days.The five chinook fry branded in Indian River which were collected in the outmigrant traps at Tal keetna Station averaged 17 days between release and recapture with a range from 8 to 26 days. A total of 9,802 chinook salmon fry were cold branded in sloughs and side channels in the middle river between July 1 and October 15.Of these fish,643 (6.6%)were later recaptured;637 in the same slough where they were originally marked and released,·seven fish in sloughs and side channels downstream from their release sites,four fish in the Talkeetna Station traps and two fish at sites upstream from their points of release.Of the 637 fry recaptured in the same slough where they were marked,136 were caught 5 to 30 days later,and 113 were caught 30-60 days later.The branded chinook fry collected in the Talkeetna outmigrant traps averaged 12 days between release and recapture with a range from 8 to 17 days. 27 CHINOOK 0+ 80 75 70 ,,-... ::E 65~......, :I: I- 0 60z W ....J Z 5S<{w ~ 50 45 40 L M;\Y E JUN L JUN 1984 E JUL L JUL E AUG L AUG E SEA..SEP-E OCT SAMPLI NG PERIOD '""" - - - ...... Figure 16.Chinook salmon (age 0+)mean length and range of lengths by sampling period for fish collected in the lower and middle reach of the Susitna River,1984. 28 - Figure 17.Weight/length relationship for juvenile chinook salmon collected at the Talkeetna stationary outmigrant traps, 1984. 29 Table 1.The number of chinook salmon fry marked and recovered in Indian River by sampling period,1984. -. Recapture Period Number Marki n9 of Fi sh July August August Sept.Sept.Oct Period Marked 16-31 1-15 16-31 1-15 16-30 1-15 Total July 1-15 2,093 26 10 5 2 3 3 49 July 16-31 1,924 5 4 5 5 2 21 -, August 1-15 6,735 8 17 8 8 41 August 16-31 3,806 4 5 2 11 September 1-15 5,492 17 7 24 September 16-30 3,356 TOTALS 23,406 26 15 17 28 38 23 147 - - - 30 -. _. - - .- - 3.1.4 Population estimates Using the mark-recapture data of Table 1 with the Schaefer method (Appendix B),there were an estimated 3,211 ,000 age 0+chinook fry in Indian River after mid July.Females comprised 41.7%of the estimated population of 8,482 (9120-7%milling)adult chinook salmon (greater than 350 mm)which passed Curry Station in 1983 [95%confidence interval (C.l.)on estimate of 9120 of 6,148 to 14,212 fish;Barrett et ale 1984J.Indian River chinook comprised 27%of the peak spawning survey counts (Barrett et ale 1984).Using a fecundity estimate of 10,622 eggs p.er female (Healy and Heard 1984),an estimated 10,143,000 eggs were deposited in Indian River during 1983.It is not possible to calculate the egg to outmigrant survival rate because of unknowns in both the adult and the fry population estimates. Population estimates were made at three sloughs and two side channels in the middle river during the cold branding study (Table 2).Populations were estimated at a high of 47,000 chinook fry in Slough 22 to a low of 3,400 in Lower Side Channel 11A.No Jolly-Seber estimate of population size was made for August 11 at Moose Slough because the head of site closed the night of August 11 and almost all of the fish left.Only one chinook fry was captured on August 12;there were no recaptures. The effect of fluctuating discharge levels on the density (beach seine catch with constant effort)and total number (population estimate).of chinook fry in sloughs and side channels can be seen in Figs.18 and 19. Estimates of population size were made using the Jolly-Seber method which allows for inmigration,recruitment,outmigration,and mortal ity. Recruitment does not occur,so all gains to the population were a result of migration into the site.Similarly,assuming that mortality during a five day period is negligible,all losses to the population were a result of migration from the site. The total number of fry in Moose Slough during these five days paralleled the density of fry and the discharge level (Fig.18).This pattern suggests that habitat qual ity was best at the highest observed flow and declined with a drop in discharge level.As the surface area of the site and the habitat quality decreased,so did the total number of fish at the site.Evidently,the site is of little rearing value to chinook salmon when the head of the site is not breached.A partial explanation is that the water clears when the head is closed;there is little cover other than turbid water at this site.The marked/unmarked ratio for each day was diluted by the entry of new fish into the site through the slough head,until the head closed.By that time,most of fish that had been at the site the previous four days had left. Residence time in this slough was low.This site probably acts mainly as an outmigration corridor and temporary rearing area. At Lower Side Channel 11A,the density of fry stayed relatively constant over the five days even though the discharge level steadily decreased (Fig.19).Meanwhile,the total number of fry at the site declined with the lowering in discharge level.The table of recaptures (Fig.19) indicates a longer residence time than at Moose Slough.This fact,and the fairly constant density,suggests that the habitat quality at this 31 Table 2.Chinook salmon fry population estimates by site for middle Susitna River sloughs and side channels and,for Indian River, 1984. Sampling Branding Recapture Estimate Population 95%Confidence Site Dates Dates Method Estimate Interval Lower Side Channel 11A 7/29 -8/1 7/30 -8/2 Schaefer 3,420 7/30 Jolly-Seber 4,962 2,466 -14,441 7/31 Jolly-Seber 1,370 1',038 -2,106 8/1 Jolly-Seber 1,245 958 -1,874 Side Channel 10 7/16 -7/19 7/17 -7/20 Schaefer 7,630 Moose Slough 8/8 -8/11 8/9 -8/12 Schaefer 4,990 8/9 Jolly-Seber 5,884 3,888 -11,141 8/10 Jolly-Seber 1.455 1.159 -2,071 Slough 22 9/8 -9/13 10/8 Petersen 47,050 39,000 -56.750 Schaefer 43.761 Slough 19 8/29 9/26 Petersen 4,550 3,200 -6.700 (.oJ N Indian River 7/1 -9/30 7/15 -10/15 Schaefer 3.211,000 •'.~)J J J ~)t J )I 1 j D ,!.~J 33 - RECAPTURE 29 DATE NO.7/30 7/31 8/1 8/2 TOTALMARK (!)22 13 6 46 .... Z 7/29 130 5 ~7/30 209 20 27 17 64 28 a:«7/31 24 26 50:lE 179 -, 8/1 173 30 30 -27 0 -5 0 0 0 0 -0 0 )(·0 .,26 :-4 ~... )(u .-.-POPULATION 0 -••·w ESTIMATE ·w.·l-.C)25 =-3 ct r:3a::0.•:E I -.ct -:I:l-I (,)0 -en t-o ·en p--- -~~- --- LIJ I 0-tBREACHING /:-2-Z ~2~-0 24 ....... ""'0 FLOW //·0 I )(·-I -/·t-/.....ct ...:I:r/·Io·...11 •••••e ·...J I (,) :-':;)r 1 I-23 a..ct·0 I Ua..I -,-t-·I I 22 0 '0 1"""'1 2 :3 4 5 DAY -Figure 19.Catch,estimated population size,and main- stem discharge 1evel at Lower Side Channel llA,July 29 -August 2,1984. - 34 ..... site is relatively.unaffected by changes in level of discharge.How- ever,the total number of fry at the site necessarily declines with a lowering discharge level because the amount of habitat (surface area) available decreases.The constant density of fry even after the head of the site closed is perhaps attributable to a greater amount of object cover at this site than at Moose Slough. 3.2 Coho Salmon 3.2.1 Catch per unit effort 3.2.1.1 Age 0+ Juvenile coho salmon were observed during the coded wire tagging study to be most abundant at Indian River.Catch rates were not recorded. The cold branding study collected 1,548 coho salmon fry in Indian River from July 1 through October 15.Of thi s catch,31%of the coho were captured in Section I,44%in section II and 26%in section III.Beach seining of sections II and III during July captured 444 juvenile coho salmon;76%in section II and 24%in section III.Minnow trapping begun in late July captured 1,129 juvenile coho salmon during 947 minnow trap days for a season average of 1.2 coho per trap day.Of these catches, 43%were recorded in the lower section,31%in the middle section,and 26%in the upper section. The catch per unit effort for all Jndi an Ri ver secti ons combi ned was steady through the season rangi ng from 1.1 to 1.5 fi sh per trap day (Fig.20).Coho fry catches were highest in section III with an average of 5.0 coho per trap day over the season.Season average CPUE in section II was 1.4 coho per trap day and Section I averaged 0.8 coho per trap day.. A total of 90 coho salmon fry were captured during the cold branding study in sloughs and side channels in the middle Susitna River.Ninety- five percent of the coho catch was recorded in slough habitats in this reach.Beach seining during July and August captured 40%of the season's total catch while minnow trapping during September and early October collected the remaining 60%(average of 0.2 coho per trap day). Daily minnow trap CPUE ranged from a low of 0.01 at Slough 22 and Side Channel 21 in September to a high of 7.6 coho per trap day at Slough 14 on September 10. Peak catches for the 1,830 age 0+coho salmon collected at the Talkeetna Station outmigrant traps were recorded during late July and August,and the highest catch rate of 2.9 coho fry per hour was recorded on July 30, by which time 50%of the season.total had been recorded (Fig.21).The last coho fry was captured in the traps on October 4. A total of 441 age 0+coho salmon were captured at the Flathorn stationary outmigrant trap during 1984.Catch rates were highest during late August and late September and the peak catch rate of 1.5 fish per hour was recorded in the trap on September 30 (Fig.22).Fifty percent of the catch at this site occurred by August 26.Only 16 age 0+coho were captured in the mobile trap at Flathorn Station. 35 - ..... -COHO JUVENILES I!\J DIAf'J RI\/ER ~I 984 8..,.---------------------------,'"'" 7 ~Section UI - -E OCT Combined...-~I L SEP ..,..,j--~--___io.I_...!ectlon I L AUG E SEP SAMPLlNG PERIOD E AUG 6 ~a ~5 «a::: l- a:::4 w 0.. ::r:3(j I- <:( (j 2 1 0 L JUL Figure 20.Coho salmon (age 0+)average catch per minnow trap by sampling period and survey section in Indian River,1984. ~, 36 TALKEETNA BOTH TRAPS COHO 0+ 100 T------------l--------:::::::=;;;=;:-i2 .3 90 SEPAUGJUL %CUMULAnVE %OF HIGHEST CPUE I l~ I~I ~ !\I I \I /'~J '\~!o -t-~~~~=:---+_----+---:::....-+_-~~~~.)...O.O 10 40 60 20 30 50 70 80 I- Z W ()a::w 0. - -- -Figure 21.Coho salmon (age 0+)smoothed daily catch per unitieffort and adjusted cumulative catch recorded at the Talkeetna stationary outrnigrant traps,May 14 through October 6,1984. 37 FLATHORN STATION AGE 0+COHO ""'" """, -.---------------------------.._1.1 ~ l:t:: ~ 0 :z: ...... % U l- e(-(,,) ~ Figure 22.Coho salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. 38 - ..." ~I ,- A tota 1 of 380 age 0+coho sa 1mon were captu red in the lower Sus i tna River during the JAHS study (see Part 2 of this report).Catch rates were highest during the late summer sampling and the peak catch rates were recorded in early October (Fig.23). The Deshka River weir captured 95 coho salmon fry during 1984;the peak catch rate of 1.3 fish per hour was recorded on July 25 (Appendix Table A-I).Minnow trap catches at this site were highest during late August at 2.6 coho per trap (Appendix Table A-2). 3.2.1.2 Age 1+and older Age 1+coho salmon were collected sporadically during the coded wire tagginq study in May and June with the highest concentrations observed in Slough 11 and Indian River.The cold branding study from July through early October captured 25 age 1+coho at Indian River and 18 at middle river slough and side channel sites during the season. Peak catches for the 1,425 age 1+coho salmon juveniles captured at the Talkeetna Station outmigrant traps were observed in mid June and were again high in late July and late August (Fig.24).Fifty percent of the catch was recorded by June 25.The highest catch rate for these age classes was 1.6 fish per hour recorded on June 18 and the last capture was on October 2. Catch rates'for the 291 age 1+coho salmon juveniles captured at the Flathorn stationary outmigrant trap were highest during late August and September (Fig.25)and the highest CPUE of 0.8 coho per hour was recorded on September 3.Fifty percent of the tota 1 ca tch was recorded by August 30 and the 1ast capture of these age cl asses was October 1. The mobile outmigrant trap captured 10 age 1+coho salmon during the season. The JAHS study in the lower ri ver coll ected 62 age 1+coho salmon juveniles with most of the captures being recorded at tributary sites in this reach. The Deshka River weir collected 26 age 1+coho while minnow trapping at this site captured 119 fish.Catches were observed throughout the season with a peak rate of 6.2 coho per trap recorded in late August. A total of 44 age 2+coho salmon juveniles were collected during the 1984 studies.Talkeetna Station,Flathorn Station,and the Deshka River accounted for 95%of the captures of this age class. 3•2•2 Growth 3.2.2.1 Age 0+ Coho fry collected in the lower river were consistently larger than the fry collected in the middle river throughout the season (Fig.26).Coho fry collected in the middle river averaged 40 mm total length during late May and showed a steady growth to a mean of 58 mm by late August. Coho fry in the lower river averaged 42 mm in early June and had grown 39 COHO.CPUE 1984 - .... - - 100 90 80 w 70 ::::J Q. ()60 l- UIw 50:I: C) :I: U.40 0 ~30 20 10 0 E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLING PERIOD ..... - Figure 23.Coho salmon juvenile catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984. 40 - .,.,.. """" ...... -- l~ TALKEETNA BOTH TRAPS COHO 1+ex:2-+ 1- Z tJ.J () 0: W CL 100 :---------r-------------=:=::;==------r 16 90 1 I' 80 l /.%ClJtviU LATIV'E I 70 1 %OF HIGHEST CPUE I :J ~! 40 ~\I :::I: 1\~ o -hAY JUN JUL ~u;hE/~O'O - Figure 24.Coho salmon (age 1+and older)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. 41 100 90 80 70 60 f- Z Wu 50 0:: Wa. 40 30 20 10 0 FLATHORN STATION COHO 1 +&2+ .....---------------------,,----'----r---r-O.6 %CUMUlATIVE %OF HIGHEST CPUE -l-__----lJ.:q:::..-+-....IL..-t-_:-:-::--_-+_-:-:::-=-__H-0.0 MAY - ""'" - Figure 25.Coho salmon (age 1+and older)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. 42 - .- COHO 0+1984 72 -,',..-----------------------------, Middle Sualtna River L AUG E SEP L SEP E OCT E JUL L JUL E AUG SAMPLING PERIOD L J·UNEJUN 56 54 52 50 48 46 44 42 40 38 -t-----r----r-----,------......---.,----.----,.-----! l MAY 70 68 66 64 62........ :::E 60~ ......,58 I I-o Z W -l Z ~ ::E Figure 26.Coho salmon (age 0+)mean length and range of lengths by sampling period for fish collected in the lower and middle reach of the Susitna River,1984. ".... 43 to a mean length of 71 mm by late September.The number of fish measured,mean length,and range of lengths by sampling period for coho fry are presented for each data collection area in Appendix Table A-6 and A-7. 3.2.2.2 Age 1+and older The average length of age 1+coho salmon juveniles collected in the lower river during the open water season was greater than that of fish of the same age class collected in the middle river (Fig.27).Age 1+ coho averaged 70 mm total length in both reaches during May and increased to 104 mm in the middle river and 111 mm in the lower river by· early October.Length data by collection area and sampl ing period are provided in Appendix Table A-8 and A-9. Age 2+coho salmon juveniles collected during the 1984 studies averaged 137.1 mm and ranged from 114 to 176 mm (Appendix Table A-10). A sample of juvenile coho salmon were measured at Talkeetna Station to provide a relationship between length and weight for fish passing this site (Fig.28). 3.2.3 Cold branding A total of 1,480 juvenile coho salmon were cold branded in Indian River from July 1 through October 15.Of these fish,five were recaptured in Indian River and two were recovered at the Talkeetna Station outmigrant traps.The marked coho recaptured in Indian River were branded and rel eased at TRM 11.5 on July 17 and recaptured at TRM 2.2 between September 9 and 11,for an average of 55 days between rel ease and recovery.The two branded coho recovered at Talkeetna Station were released in Indian River on August 12 and were recovered in the outmi- grant traps on August 31 and September 22;19 days and 41 days, respectively,between release and recovery. A total of 106 juvenile coho salmon were cold branded at slough and side channel sites,and the only recapture was recorded at Talkeetna Station. The recaptured fish was marked and released at Slough 14 on September 10 and was recovered in the traps on September 16. 3.2.4 Population estimates Since only 100 to 200 of the estimated 750 adult coho passing Curry Station in 1983 entered Indian River,and since juvenile coho of the same brood year outmigrate as age 0+,1+,and 2+fish,few juvenile coho salmon were captured for marking during the 1984 cold branding studies. Too few branded coho salmon were recaptured to provide population estimates for any of the sites surveyed. 3.3 Sockeye Salmon 3.3.1 Catch per unit effort 44 ~, - - - COHO 1+1984 115 110 ,,// 105 // /,-100 ff........ :i :::iE 95.......//::I: I- 0 90 ///zw /...l z 85 «:(w :::iE 80 75 ,,-.70 65 MAY JUNE JULY AUG SEP-OCT SAMPLING PERIOD Figure 27.Coho salmon (age 1+)mean length by month for fish collected .-.in the lower and middle reach of the Susitna River,1984. 45 COHO SALMON ~, ~, 10g e y =-12.21 +3.12 loge x r 2 =O.98 22 20 18 16 14 E Cl' 12 4-'.s::. ())10'4) 3:..,8 ~ 6 4 2 0 40 60 80 Total length (mm) 100 o 120 - Figure 28.Weight/length relationship for juvenile coho salmon collected at the Talkeetna stationary outmigrant traps, 1984. 46 - !~ ,...,., I 3.3.1.1 Age 0+ Sockeye salmon fry were collected during the coded wire tagging study in May and June at sloughs 8A,9,11,and 21 but catch rates were recorded only for Slough 21.These data were determi ned from 24 hour fyke net catches and are presented in Appendix Table A-11. A total of 248 sockeye salmon 'fry were captured at slough and side channel sites in the middle river and in Indian River during beach seine sampling conducted in July and August.Of these fish,94%were col- lected in sloughs and the remaining 6%were collected in Indian River and at mainstem side channels. Peak catch rates for the 7,484 age 0+sockeye salmon fry collected at the Talkeetna Station outmigrant traps were recorded in mid June and early July with the highest da"ily catch rate of 13.0 sockeye fry per hour occurring on June 18 (Fig.29).The major downstream redistri- bution of sockeye fry in this reach had occurred by mid July (50%by July 4).The last sockeye fry at Talkeetna Station was observed on October 4. Juvenile sockeye catches at the Flathorn stationary outmigrant trap were greatest during May and June but the downstream movement of sockeye fry continued through the open water season (Fig.30).A total of 2,315 sockeye fry were collected in the trap during 1984,and the peak catch rate of 4.6 fish per hour was recorded on June 8.Fifty percent of the catches had occurred by June 29 and the last capture was October 1. Mobile trap catches of sockeye fry at Fl athorn Stati on were hi ghest during June and the peak catch rate of 5.4 fish per hour was recorded on July 12 (Fig.31).Of the 114 sockeye collected in the mobile trap during 1984,59%were captured at bank transect points (Fig.32). A total of 412 sockeye salmon fry were collected in the lower river during JAHS surveys from June through.October (see Part 2 of this report).Catch rates at JAHS sites peaked in late June and then were low throughout the remainder of the season (Fi g.33).An increase in catch rates was recorded at some sites including Rolly Creek'(RM 39.0) and Beaver Dam Slough (RM 86.3)in late August and September,indicating the movement of sockeye int,o these sites during late summer. 47 - ""'Ii T)A,LK~EThJA BOTH TRAPS SOCKEYE {~"+\_.1 100 1 12•7 -I r- gO 1 I f l- 80 ! (I 70 I l %CUMULATIVE ~~ !%OF HIGHEST CPUE I a:60 r~I-\0z ~w ~%(j SO !......0:: W /i::z:::Q.\(.)40 1-1-~, \ I « 30 ~(.) ~-~-20 Il-I 10 Ir ~ ~0 , 0.0 JUN JUL AUG SEP Figure 29.Sockeye salmon (age 0+)smoothed daily catch per unit effort .and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. ~, 48 FLATHORN ST.ATION SOCKEYE 0+ 100 90 80 70 60 ~ Z I.&J <J 50 0:: I.&J Q. 40 30 20 10 0 T--r;--------------=::;:::::::::::===:::;;::;:~13.5 %CUMU LATI\iE %OF HIGHEST CPI)E A/"--J-F-----il-------t------+------+-~~-~-+-0.0 "'UN JUL AUG SEP Figure 30.Sockeye salmon (age 0+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. 49 - - - - / - - - -- , -",If IF",v 1 r IflRRaabaaaa••V ~b a a - 100 90 l.i.l 80 .~. CL e..;.70 I- [J) W :r:60 D I: l.L 50 0 LLI 13 4-0<I- Zw ,30 1:-'1 0::: W Q..:20 10 0 1215 20 JULY .30 1 5 DATE -----------1 a -Not Sampled I b -No FI8h Captured I a~ba ~bb baababb~-rT-rr-J- 10 AUGUST 20 2530 Figure 31.Sockeye salmon (age 0+)daily catch per unit effort recorded at the Flathorn mobile outmigrant trap,July 12 through August 31,1984. 50 ~I' AGE 0+SOCKEYE SEASON CATCH 24 22 20 :J:18 0.....«16<J ..J 14~ 0....12 l.L. 0....10z W <J 8a::wa.6 4 2 0 "'1 2 3 4 "'5 *6 "'7 8 9 "'10 TRANSECT POINT NUMBER '"BANK TRANSECTS Figure 32.Sockeye salmon (age 0+)percent of the total catch by sampl ;ng po;ntr :recorded at the Fl athorn mob;1e outmigrant trap,1984. 51 90 1 00 -""---"""T?""~""""------------------------' w ~ Q..o tiiw:r:(..=l :J: 1.La ~ 80 70 60 .so 40 30 20 10 SOCKEYE CPUE 1984 ~- - "'"" E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLI NG PERIOD Figure 33.Sockeye salmon juveniles catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984. 52 TALKEETNA &FLATHORN SOCKEYE 1 + . 100 Tr-----::::=::::==::;;=::;::::;:r----------i SEPAUG %CUMULATI\IE %OF HIGHEST CPUE 70 10 40 20 60 30 80 O.....f+'----+------'~--J...I.---L.\.-....u...f__----_t_-----_H 90 ....zw.()50a::wa. Figure 34.Sockeye salmon (age 1+)smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn and Talkeetna stationary outmigrant traps,May 14 through October 6,1984.. - - 53 ~,'3.3.2 Growth The mean length and range of lengths for age 0+sockeye salmon by reach of river and sampling period is presented in Fig.35.During May and June,sockeye fry collected in the middle river reach had a smaller mean length than the same age class sockeye collected in the lower river.By early July,sockeye fry averaged the same length (49 mm)in both - reaches,and by late August,middle river sockeye fry were averaging larger than fish collected in the lower river.This trend continued through the remainder of the season.The number of fi sh measured,the mean length and range of lengths by sampling period for sockeye saJmon . fry are presented for each of the data collection areas in Appendix Table A-12. '""'" The 90 age 1+sockeye salmon collected during 1984 ranged from 56 to 102 mm total length (Appendix Table A-13).A coded wire tagged sockeye fry released in 1983 and recaptured in 1984 had increased from 32 mm to ~ 81 mm. A sampl e of juveni 1e sockeye were measured at Talkeetna Station to provide a relationship between length and weight for fish passing this site (Fig.36). 3.3.3 Coded wire tagging and recovery A total of 14,532 tagged sockeye salmon fry averaging 33 mm total length were released between May 22 and June 22,1984 (Table 3).Tag retention ~ rates for sockeye fry averaged 97.1%and ranged from 92.3 to 99.0%. Tagging mortality ranged from 0.6 to 2.6%and averaged 1.3%. ..... A total of 366 tagged sockeye salmon fry (2.5%of the total tagged sockeye released)were recovered from the 7,484 age 0+sockeye captured and examined for tags at the Talkeetna Station outmigrant traps during 1984.In addition,15 sockeye fry with clipped adipose fins but no coded wire tags were recovered in the traps.When compared to the total tagged sockeye salmon fry recovered,this provides a tag retention rate at the traps of 96.1%. Trap recoveries of coded wire tagged sockeye fry were made from a to 109 days (mean =35 days)following their release at the tagging sites (Fig. 37).In addition,one tagged sockeye fry which was released from Slough 21 on May 28 was recaptured at Flathorn Station on July 7.Seven coded wire tagged sockeye fry were recovered during the cold branding study in early August (Tabl e 4).Six of these fi sh were recovered at Moose Slough (RM 123.2)and one tagged sockeye fry was recovered at a side channel below Slough 11 (RM 135.2). A single coded wire tagged sockeye salmon marked and released during 1983 was recovered during the 1984 sampling season.This fish was released June 8,1983 at Slough 11 and was recovered at Talkeetna Station on July 21,1984. _. 54 ,- SOCKEYE 0+"1984 L AUG E S EP L S EP E OCT L JUL E AUG SAMPLING PERIOD E JULLJUNEJUN 52 50 48" 46 44 42 40 38 36 34 .32 30 -+---.......---,---,.-----.---.......----,,.-----.-----1 L I'v1AY 62 .,--------------------------..., 60 58 56 54 Figure 35.Sockeye salmon (age 0+)mean length and range of lengths by sampl ing period for fi sh call ected in the lower and middle reach of the Susitna River,1984. 55 SOCKEYE SALMON 6-r------------------------------. 10Qe y=-12.33+3.IIIOQ.K r 2 =O.97 4 1 5 ~ E 0' 3 1If%~ 4-' .J:en 'tj ~ tJ 2>:J 907050 O-l----.,.----,...-----,~--___,.---__r---_r_----1 30 Total length (mm) Figure 36.Weight/length relationship for juvenile sockeye salmon collected at the Talkeetna stationary outmigrant traps, 1984. 56 Table 3.Coded wire tag release data for sockeye salmon fry on the Susitna River by tagging site and release date,1984. Tagging Site Number of Date of Percent Tag Percent (River Mil e)Fish Tagged Release Retention Mortality Slough 21 3,736 5/28 97.9 2.6a (RM 142.0) Slough 11 2,327 5/22 92.3 1 .1 (RM 135.3)2,732 5/24 97.7 0.7 1,537 6/22 96.6 1.1-Slough 9 2,052 6/9 99.0 1.0 (RM 128.3) Slough 8A 2,148 6/19 99.0 0.6 (RM 125.3) TOTAL -ALL SITES 14,532 5/22-6/22 97.1 1.3 a Mortality due to handling,thermal,and anesthetic stresses. 57 CODED WIRE TAGGED SOCKEYE 80 70 I- ::I: C)60~ (J ::I: (J)50 ii: Clw 40C) 0<t: l- lL.300 0:: W CD 20~ ::::l Z 10 0 5 20 35 50 65 80 95 110 NUMBER OF DAYS AFfER RELEASE (Grouped by 5 Day Period) Figure 37.Length of time between the mark and recapture of coded wire tagged sockeye salmon juveniles in the middle reach of the Susftna River,1984. 58 ,~ - - ..!, Table 4•Recoveries of coded wire tagged sockeye salmon fry at mainstem river sites between Tal keetna and Devil Canyon, 1984 • ....Collection.Collection Release Release Site Date Site Date Moose Slough 1 8/8 Slough 21 5/28 r-Moose Slough 8/8 Slough 21 5/28 Moose Slough 8/8 Slough 11 6/22 Moose Slough 8/8 Slough 9 6/9 Moose Slough 8/8 Slough 8A 6/19 Moose Slough 8/8 Slough 8A 6/19 Slough 11 Side Channe1 2 8/3 Slough 21 5/28 1 River Mile 123.2~ 2 River Mile 134.9 59 The ratio of coded wire tagged sockeye fry to total sockeye fry was the same (0.05:1.00)in both traps at Talkeetna Station.This indicates that the coded wire tagged fish were uniformly mixed in the total population by the time they migrated past the traps. 3.3.4 Population estimates and survival rates of outmigrants Females comprised 38.5%of the population of 1,900 adult sockeye salmon estimated past Curry Station in 1983 (95%C.!.-1,600 to 2,300 adu1 ts) and the fecundity of Susitna River sockeye averaged 3,350 eggs per .female,with a 95%C.I.of 3131 to 3569 (Barrett et al.1984).Milling activity was estimated at 30%(Barrett 1984).These data provided a calculation of total potential egg deposition for sockeye salmon of 1,715,000 eggs during 1983. Using the method outlined by Schaefer (1951),the number of age 0+ sockeye salmon fry above Ta"lkeetna Station during 1984 was estimated to be 299,000 (Appendix Table B-1 and B-2).A comparison of this estimate to the calculated potential egg deposition (dividing the estimated number of fry by the number of eggs)gave an egg-to-outmigrant fry survival rate of 17%.The reliability of this estimate is not currently known because there is no way to estimate the variance of the adult mi 11 i ng estimate and because we do not currently have a method of estimating the variance on the Schaefer estimate of the fry population size. 3.4 Chum Salmon 3.4.1 Catch per unit effort Chum salmon were collected during the coded wire tagging study in May and June and during beach seine sampling of Indian River in July.Catch rates were not generally recorded during these studies except for 24 hour fyke net sets at Slough 21 (Appendix Table A-10). Peak catches of chum fry collected at the Talkeetna Station outmigrant traps were recorded during late May and mid June,with the highest daily catch rate of 8.0 fish per hour occurring on June 14 (Fig.38).Ninety- fi ve percent of the 3,590 chum fry captured at Tal keetna Stati on were recorded by July 15.The major outmigration had occurred by the end of June (50%by June 13),although the migration continued until September 11. Chum salmon fry catches at Flathorn Station were greatest during June with a peak catch rate of 10.9 fish per hour recorded on June 14 by which time 50%of the season catch had occurred (Fig.39).By July 1, 97%of the chum fry collected at this site had been captured;the last chum fry was captured on July 22. Beach seining and electrofishing at side channel,slough,and tributary sites in the lower river reach collected chum salmon fry during June and July (see Part 2 of this report).Chum fry were abundant in this reach during early June but catches steadily decreased through July (Fig.40). 60 - -~ - - ~ T,A.Lk~EFTI',JA BOTH TR.A.PS CHUM FR"'( 100 -[ t·· 2~..... /9°1 \so -I 70 ~\%CUMU LATI\/E .I ~~I ~\I %OF HIGHEST CPUE 80 _I (~I~-zw 0 50 -10::-0 W t:i...~ 40-et U Ii r I - 30 -./~~I"'"" 2')-1'1/1'I V\j~-l:V ~) I JUL ~I 0.0 MAY I JUN AUG SEP ~ ...., Figure 38.Chum salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. - 61 ~, ~ FLATHORN STATION CHUM FRY 100 5.8 - 90 l -~80 70 60 %CUMULAnYE l~l-%OF HIGHEST CPUEzw ~ ()50 0:I :r:w I <.iIi. 40 I- <l:-u 30 20 10 0 0.0JUNJULAUGSEP ~, Figure 39.Chum salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant trap,May 20 through October 1,1984. - 62 r-. CHUM CPUE ·1984 100 -rr~;o-r------------------------..., 90 .- - .- I W :::Ja.u t- (.I! W:::c Cl :::c lJ...o te 80 70 60 50 40 30 20 10 E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLING PERIOD - Figure 40.Chum salmon fry catch per unit effort by sampling period recorded at JAHS sites in the lower reach of the Susitna River,1984. 63 3.4.2 Growth At both Talkeetna and Flathorn Stations,chum length ranged from emer- gent lengths (less than 35 mm)to lengths greater than 60 mm for May, June,and July (Appendix Table A-14).Chum salmon spawn in both tribu- taries and sloughs and there is a wide range in emergence timing.The fish caught at 30-40 mm are probably recent emergents.The 50-60+mm fish have gained over 20 mm in length. During June,Indian River chum fry averaged 40 mm and had increased to a mean length of 48 mm by early July.Limited sampling of the Talkeetna River during June and July indicated a mean length of 43 mm for chum fry outmigrating from this tributary. 3.4.3 Coded wire tagging and recovery A total of 31,396 tagged chum fry averaging 43 mm total length were released between May 22 and June 22,1984 (Table 5).Tag retention rates ranged from 93.0 to 100%and averaged 96.4%.Mortality rates between tagging and release averaged 0.9%and ranged from 0.0 to 2.7%. Fifty-one tagged chum salmon fry (0.2%of the total tagged chum released)were recovered from the 3,590 chum salmon fry captured and examined for tags at the Talkeetna Station outmigrant traps during 1984. In addition,two chum fry with clipped adipose fins but no coded wire tags were recovered in the traps.When compared to the total tagged chum salmon fry recovered,this provides a tag retention rate at the traps of 96.2%. Trap recoveries of tagged chum fry were made from 0 to 29 days (mean =8 days)following their release at the tagging sites (Fig.41). The ratio of coded wire tagged chum fry to the total number of fish caught at each trap at Talkeetna Sta ti on was 0.016:1 at Trap 1 and 0.013:1 at Trap 2,indicating that the tagged chum fry were randomly distributed with the untagged population by the time they migrated past the traps. 3.4.4 Population estimates and survival rates of outmigrants Adult population estimates at Curry Station during 1983 were 21,100 chum salmon with 95%confidence limits of 19,200 to 23,500 adults.Females comprised 34.5%of these fish and chum salmon milling was estimated at 40%(Barrett et ale 1984).Fecundity of Susitna River chum salmon was determined during 1983 to be 2,850 eggs per female (95%confidence limits of 2,666 to 3,034).These data provided an estimated total potential egg deposition of 12,448,000 eggs. The population estimated using the Schaefer (1951)method was 2,039,000 chum salmon fry outmigrating past Talkeetna Station during 1984 (Appen- dix Table B-3 and 8-4).Using the above data,an egg-to-outmigrant fry survival rate of 16%was calculated for chum salmon.As with sockeye salmon,there is no way of knowing the reliability of the estimate 64 - - - - """': Table 5.Coded wire tag release data for chum salmon fry on the Susitna River by tagging site and release date,1984. Tagging Site Number of Date of Percent Tag Percent (River Mile)Fish Tagged Release Retention Mortality Slough 22 2,383 6/1 98.0 0.5 (RM 144.3) r- Slough 21 2~201 6/3 96.6 1.4 (RM 142.0) Slough 20 1,255 6/11 96.9 0.6 (RM 140.1) Slough 15 351 6/14 100.0 0.0 (RM 137.3) Indian River 4~612 6/1 94.5 0.7 (RM 138.6)341 6/1 93.0 O.Oa 4~592 6/21 93.8 2.7 -"2,511 6/22 95.0 0.4 Slough 11 2~031 5/22 97.7 0.1 (RM 135.3)2~203 5/24 93.9 0.3 572 5/24 99.0 0.2 1~916 6/16 98.0 0.4 Slough 9 5~122 6/6 99.4 0.7 (RM 128.3) Slough 86 1~306 6/13 98.0 0.8 (RM 122.4) ,... TOTAL -All SITES 31~396 5/22-6/22 96.4 0.9 a High mortality due to injury from improper headmold. - - .- 65 CODED WIRE TAGGED CHUM'SALMON -27246 9 12 15 18 21 NUMBER OF DAYS AFTER RELEASE 3o 15 ""'T"""---------------------------, 14 13 12 11 10 9 8 7 6 5 4 3 2 1 O-"r-l..I!r.l-"r-l.l~~f.J_l',,..J_I',..u:,..L.r__.___.__.,..._,....._I!,...l..lIT.uyuyJiT_U~T.L..r__I'r...l.....r...y.....,......__._..,_JlT-UY .......:r.: <=J ~ (J .:r.: CJI Li: o I.LI <=Jo ;! La..o 0:: I.LI CD ~ :::J Z Figure 41.Length of time between the mark and recapture of coded wire tagged chum salmon juveniles in the middle reach of the Susitna River,1984. 66 - ,I"'- . 1"- ,.- because the variance of the adult milling estimate and the variance of the fry population estimate are not known. 3.5 Pink Salmon Sixty-eight pink salmon fry were captured between May 15 and July 18 at the Talkeetna Station outmigrant traps during 1984,with the peak catch rate of 0.8 fish per hour being recorded on June 18 (Fig.42).Pink fry migrating past Talkeetna Station averaged 36 mm total length with a range from 29 to 53 mm. A total of 405 pink salmon fry were collected in the stationary outmi- grant trap at Fl athorn Station.Catches occurred from May 21 through July 6 and the peak catch rate of 4.0 fish per hour was recorded on June 5 (Fig.43).Fifty percent of the catches at this site were recorded by June 11.Pink fry collected at Flathorn Station averaged 34 mm and ranged in length from 25 to 46 mm. No pink salmon fry were collected during the cold branding studies in the middle river,during sampling of the Deshka River,or at JAHS sites in the lower river during 1984 • 3.6 Descriptive Statistics for Catch and Environmental Variables Summary statistics for Talkeetna Station catch are given in Table 6 and for environmental variables in Table 7.Flathorn data are summarized in Table 8.The influence of discharge peaks on the level of outmigration can be seen by comparing the seasonal discharge level (Fig.44;Fig.45) with the outmigration plots presented earlier.Results of a statistical time series analysis of 1983 and 1984 discharge,turbidity,and age 0+ chinook and sockeye salmon outmigration are presented in Appendix C. 67 TALKEETNA BOTH 'TRAPS PI i'.n< - 100 0.2 !I 90 I !80 I I 70 1%CUMULATIVE a:: 60 I I Jroo%OF HIGHEST CPUE 0z LU 1 ::c ()50 t-,0:: W I :t:n.j l)40 I- et l) 30 10 o ..pC-.i.......L---L..lf----U.+-...l-I.-L---"--_-++-+-_+_.0.0 AUG SEP Figure 42.Pink salmon fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,May 14 through October 6,1984. 68 '""" - - - ..- - Fi gure 43.Pink sal men fry smoothed daily catch per unit effort and adjusted cumulative catch recorded at the Flathorn stationary outmigrant traps,May 20 through October 1,1984. 69 Table 6.Summary statistics for juvenile salmon catch per hour by species and age cl ass recorded at the Tal keetna Station outmigrant traps,May 14 through October 6,1984. Catch Per Hour,Both Trapsa - Chinook 0+ Chinook 1+ Coho 0+ Coho l+b Sockeye 0+ Sockeye 1+ Chum Min 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Max 17.2 3.5 2.9 1.7 13.0 0.3 8.0 Mean 2.2 0.3 0.3 0.3 1.2 0.0 0.7 Std.Dev. 3.2 0.6 0.4 0.3 1.8 0.0 1.2 a n =146 b includes all juvenile coho age 1+or older. ...., Table 7.Summary stati.stics for habitat variables recorded on the Susitna River between the Chul itna River confl uence and Devil Canyon,May 14 through October 6,1984. """i Min -Max Mean Std.Dev.n Discharge (ft3 jsec)a 6,780 52,000 19,405 8160.0 146 Water Temperature (oC)b 2.0 13.5 8.8 3.0 145 Turbidity (NTU)b 13 400 115 92.0 145 a USGS provisional data at Gold Creek,1984. b ADF&G data at Talkeetna Station outmigrant traps,1984. 70 - - - Table 8.Summary statistics for juvenile salmon catch per hour by species and age class recorded at the Flathorn Station outmigrant traps,May 20 through October 1,1984 . ..- 'Catch Per Hour a Min Max Mean Std.Dev. Chinook 0+0.0 7.8 0.7 1.1 Chinook 1+0.0 6.5 0.1 0.6 Coho 0+0.0 1.5 0.1 0.3 Coho l+b 0.0 0.8 0.1 0.1 Sockeye 0+0.0 4.6 0.8 0.8 Sockeye 1+0.0 0.4 0.0 0.1 ~Chum 0.0 10.9 0.3 1.1 Pink 0.0 4.0 0.2 0.5 Discharge {ft 3 /sec)c 40,800 166,000,93,122 28,887.5 .-a n =134. b Includes all juvenile coho age 1+or older. /""" c USGS provisional data at Susitna Station,1984. - -71 - ,- IS'"U 0-~- W 1121~ ::) ~.".,« 51313~ W S -0- 4121121 :;:)-L I-W Z~-121 31210 >- I- 21313 H6eeeeCl H--seeee 1130 enTurbidity~en :::>-0 4eeee a I-- W (!)3eeee Q:: <C::c 2eeeeu (J) H leee00 i0 172431 7 MAY JUN 14 21 28 5 12 19 26 2 9 JUL AUG 1984 16 23 30 6 I 3 213 27 4 SEP OCT - Figure 44.Mainstem discharge,water temperature,and turbidity in the middle reach of the Susitna River,1984.Discharge was measured at the USGS gaging station at Gold Creek.Water temperature and turbidity ~ were measured at Talkeetna Station. - 72 -- - ,.....-OJ 16eeee-o-W .281/188 CJ ct<~ (J f1J-C - - II 17 14 111 7 '4 II ••.1 I'•11 I .1 • •I .1 •117 4 MA Y JUNE JULY AUGUST SEPTEMBER OCTOBER 1984 Figure 45.Mainstem discharge in the lower reach of the Susitna River measured at the USGS gaging station at Susitna Station,1984. 73 4.0 DISCUSSION 4.1 Chinook Salmon 4.1.1 Outmigration Fifty percent of the outmigration of age 0+chinook salmon past Talkeetna Station during both 1983 and 1984 had occurred by mid July, but the rates and timing were different between the two years (Fig.46). Duri ng 1983,two pul ses of ch"j nook fry movement were recorded,one in late June.and the second in mid August.Conversely,the 1984 out- migration did not start until mid June and was then relatively steady through late August. Low tributary flows during July of 1983 trapped chinook fry in pools and side channels in Indian River until high tributary flows from heavy rainfall in mid August allowed access or flushed fry to the Susitna Ri ver (Roth et a 1.1984).In 1984,mi nnow trap catches of marked and unmarked chinook in Indian River during the cold branding study showed the movement of chinook fry out of this tributary continued from July through early October. In 1984,age 0+chinook salmon in the middle river that had outmigrated from the tributaries were found predominately in shallow,turbid,rocky bottom areas in breached sloughs and side channels during July and Augus t.Not unt i 1 mi d August,when rna i nstem flows had decreased and many of these sloughs and si de channel s were no longer breached,did catches of juvenile chinook increase at clear water sloughs and side channels.In early September,juvenile chinook were concentrated at the mouths of clearwater sloughs and side channels,but as water tempera- tures and stage continued dropping through September and early October, these fish slowly dispersed within these sites with the major concen- trations being found in areas with non-imbedded substrate and a groundwater source. The rates of outmigration of age 1+chinook salmon past Talkeetna Station were similar in 1983 and 1984 (Fig.47),but the date by which half of the total seasonal outmigration occurred was ten days earlier in 1983 than in 1984,primarily because of the late start of outmigration in 1984. The chinook fry appear to associate with the banks of the river during their downstream movement.Although juvenile chinook were captured across the entire river at Flathorn Station,60%of the total mobile trap captures were recorded at bank transect sites. 4.1.2 Freshwater life history Chinook salmon juveniles in the middle river appear to group into three separate categories.The first group are those juveniles which rear and overwinter in their natal tributaries and outmigrate to the ocean as age 1+fi sh duri n9 the spri ng of thei r second yea r.The second group of chinook juveniles spend a portion of their first summer in their natal tributaries and then,probably because of density dependent interaction, 74 - - - ~ I j. --, - - - ..... ,~ -- 1983 & 100 90 80 l1J 70>3 60 :::J :E :::J 50Q t- Z l1J 40Qa:: l1Ja..30 20 10 0 MAY 1984 TALI'<EETNA CHINOOK 0+ - .- Figure 46.Chinook salmon (age 0+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. 75 1984 TALKEETNA CHII'JOOK-1983 & 100 90 80 w 70>i= :5 60:J ~ ::::J 50() I- Zw 400 0:: Wa.30 20 10 0 JUL AUG l+ ~- - - Figure 47.Chinook salmon (age 1+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. 76 - ,.... behavioral changes related to downstream re-distribution~or flushing by high flows~enter the mainstem river.These fish actively search out suitable habitats as they move downstream.Many of the fish enter sloughs and side channels in the middle river to overwinter while others continue downstream to the lower river.Since 80%of the Talkeetna Station trap catch had occurred by August 1~and high catches were still being recorded at Indian River and selected sloughs.above Talkeetna Station in August~September,and October~it appeared that a signi- ficant percentage of 1983 brood year chinook salmon belonged to one of these two groups.We do not know what this percentage was. A third group of chinook salmon juveniles may be present in the Susitna River.Data collected at the Flathorn Station outmigrant trap showed that a porti on of the age 0+chinook were movi ng downstream past thi s site.Many o-f these were probably fry from the Deshka River.Although it is possible that these fish overwintered in freshwater habitats below Flathorn Station~it appeared that many of these fish entered the ocean as age 0+fish because few rearing chinook fry were found at sites below the Deshka River during 1984 (see Part 2 of this report). Intermittent operation of an outmigrant weir on the Deshka River during 1984 showed that a large number of age 0+chinook fry were outmigrating from this tributary during July and August.Similar data were collected in 1980 by Delaney et ale (1981),who postulated that the observed outmigration was a size related response as the fish reached approxi- mately 80 mm.It is not known whether these fish remain in habitats associated with the mainstem river or if they continue to the ocean as age 0+fish. Scale samples collected from returning adults at Sunshine Station and above indicated that the age 0+class of outmigrants represented less than 3%of the middle river returning chinook during 1983 (Barrett et ale 1984)and less than 1%in 1984 (Barrett et ale 1985).However,no adult chinook scale samples were taken in 1984 at Flathorn Station, which did not begin operation until early July.It may be that a significant proportion of the adults bound for lower river tributaries such as the Deshka did outmigrate during their first summer. Otherwise~if it is assumed that a significant percentage of Susitna River chinook salmon migrate to the ocean as age·0+fish~then either the marine survival of this age class is very low or the freshwater life histories on adult scales were not interpreted correctly.Ri chards (1979)reported that a major portion (72%)of the adult scales analyzed from the Deshka River during 1978 indicated that the fish had migrated to the ocean during their first summer as age 0+fish.Scale analysis from creel census samples collected in the Deshka River have classed these fish as predominantly age 1+outmigrants (Kubik 1967;Kubik and Wadman 1978;Kubik and Delaney 1980).. There are many unanswered questions about chinook fry life history in the Susitna River.Aging of adult chinook at Flathorn Station during 1985 will help answer the question of whether there is a significant proportion of returning adults which outmigrated during their first 77 summer.However,we still do not know the proportion of returning adul ts whi ch,as fry,followed one or the other of the three 1i fe history strategies discussed above.The answer to this question is of major importance in assessing dam-related effects on the population. 4.1.3 Estimates of population size and residence time The Schaefer population estimate of 3.2 million chinook salmon juveniles in Indian River in 1984 must be qualified.A successful method of sampling large numbers of juvenile chinook and a location containing large numbers were not found unti 1 mj d July,at whi ch time over 50%of the Tal keetna Station trap catch of age 0+chinook fry had occurred. Therefore,this estimate is only for those fish in Indian River for the period July 15 to Oct.15. The efficiency of minnow traps decreases when flows are high.Because the marked fish were not randomly re-introduced into the system,we have to assume that the.recapture was random.However,there is some reason to believe that the unmarked fish were more l'ikely to redistribute downstream during high flows than were the marked fish,which were re-introduced into side sloughs. Having two separate groups of juvenile chinook within Indian River, those fish which overwinter in Indian river and the middle Susitna River and those fish which migrate out of this reach,further complicates the population estimate.Most marked fish were marked near the mouth of Indian;it is likely that fish captured near the mouth were going to migrate out of Indian River during the first summer.Also,it has to be assumed that these fish,when transported back upriver,randomly mixed with the other fry.The estimate of 3.2 million fry for Indian River should be used as a rough approximation,obtained by an experimental project.Information gathered during the 1984 season will enable a more refined estimate for the 1985 season. The chinook fry population estimates made for sloughs and side channels give a general idea of how many fry these sites can support.The day-to-day variation in total number of fish at these sites,which results from variation in discharge level,is striking.Another impor- tant result of this study is the residence time of rearing chinook fry at these sites because of the implications this has on the results of the IFIM and RJHAB models of rearing habitat {presented in Part 2 of this report}.Habitat value from the models is measured by weighted usable area {WUA},which depends only on water depth,water velocity, cover,and substrate.The model will predict discharge levels at which habitat value of a site is high.However,there may not be many fish at a site,even when WUA is high,because of previous flushing of the site· by a high discharge or because of a seasonal effect in level of out- migration.More importantly,if the fish are using a site only as an outmigration corridor,as appeared to be the case at Moose Slough in mid August,then it really doesn't matter if the WUA is high or low,because WUA measures only rearing habitat quality.On the other hand,if the fish have a longer residence time at a site,such as at Lower Side Channel llA in late July,then the amount of WUA is important. 78 - ..... - t~ - ...... Of the 643 chinook fry which were captured in a slough or side channel, cold-branded,and later recaptured at the same site,113 were still present 30-60 days later.This indicates that a substantial amount of chinook fry rearing occurs at these sites. 4.1.4 Growth The increase in mean length of age 0+chinook by sampling period for the combined data collected at the Talkeetna Station outmigrant traps during 1982,1983 and 1984 is presented in Fig.48.Chinook fry,which emerge from the gravel at an average length of approximately 37 mm,had increased to an average of 44 mm by early June.By the end of the open-water season,their mean length was 63 mm.Chinook fry collected in the lower river in 1984 averaged from two to ten rnm larger than their counterparts in the middle river through the season (Fig.16). Chinook fry which overwinter in Indian River show little growth between late October (when they are a little less than 70 rnm long)and late March (ADF&G,unpublished data).Outmigrating age 1+fish at Talkeetna station averaged 90 mm during the peak of outmigration,so they had grown about 20 mm during April,May,and June. Examination of the downstream redistribution of juvenile chinook salmon in'the Susitna River by age class during 1984 shows that chinook fry in the middle river averaged approximately the same length (50 to 55 mm) throughout the period of peak outmigration (late June through early August).This results in very little separation between cumulative movements recorded for catch and biomass at Talkeetna Station (Fig.49). The outmigration of chinook fry in the middle river appears to be triggered,in part,by the fish reaching a critical size.As they reach this critical size (estimated at 55 mm),chinook fry redistribute down- stream to other rearing areas. In the lower river,total biomass movements were delayed in comparison to the total number of chinook fry moving past Flathorn Station (Fig. 49).This was due to the growth occurring in the lower river and because of the mixed stocks present in this reach. 4.2 Coho Salmon 4.2.1 Outmigration The downstream movement of coho salmon fry past Ta 1 keetna Stati on is compared for 1983 and 1984 in Fig.50.Although the outmi grati on from May through early July was slower during 1984,50%of the total season outmigration was recorded ten days earlier in 1984 than in 1983.The delay in downstream movement observed during July of 1983 was due in part to low tributary water levels during this period,and the high rates of downstream movement recorded in mi d August corresponded to a period of heavy rainfall and high tributary discharges • 79 CHINOOK 0+MEAN LENGTH 80 75 .-70 :::::E :::::E......., ::I:65I- Cl Z W ....J 60 ....J <{ I-a 55I- z ~ ~50 45 <)Maxlmum Combined +Minimum E SEP L SEP OCT L AUG +4O-j-----'T----r------r-----r---..-----,------r-----l L MAY E JUNE L JUNE E JULY L JULY E AUG SAMPU NG PERIOD Figure 48.Chinook salmon (age 0+)mean length and range of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984. ~, - 80 1 )J J J )))))J I 1 J J 1 ~ ::l ::i ::lo ~wo '"~ '1984 TALKEETNA CHINOOK 0+ 100 •:::;;JF='"fB II I 90 80 70 60 50 40 30 20 10 o •{:Ii'I I I i I L MAY E JUNE L JUNE E JULY L JULY E Aue L AUG E SEP l SEP E OCl SAMPLING PERIOD 1984 FLATHORN CHINOOK 0+ 1984 TALKEETNA CHINOOK 1 + 100 90 80 ~70 ~ ::l 60.::Ii ::l 0 !i;;50 w 0 '"40w Q. 30 20 10 T ,.-.-I I I 1 ,,I L M!W E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP l SEP E OCT SAMPLING PERIOD 1984 FLATHORN CHINOOK 1 + co I-' w ?; ~ ::l ::i ::lo...z Wo '"W Q. 100 I :;)i T II I 90 80 70 60 50 40 30 20 10 01 If --r-I I,I I I I L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP l SEP E OCT SAMPLING PERIOD ~ ::l ::i ::lo...z ~ W Q. .100 1 ~II I 90 80 70 60 50 40 30 20 10 o T'i I I Iii I i I L MAY E JUNE L JUNE E JULY L JULY E Aue LAue E SEP L SEP E OCT SAMPLING PERIOD Figure 49.Chinook salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and Flathorn stations,1984. 1983 &1984 TALKEETNA COHO 0+ 100 90 80 w 70>;:= :5 60 :=l 2 :J 500 I- Zw 400 0:: Wa..30 20 10 0 MAY Figure 50.Coho salmon (age O+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. 82 - ~, - -- - - - - The downstream movement of age 1+coho salmon past Talkeetna Station was approximately two weeks later in 1984 than in 1983 while the rates of movement were fairly stabl~throughout both seasons (Fig.51). 4.2.2 Freshwater life history Most coho salmon juveniles spend one or more years in the Susitna River before migrating to the ocean.Analysis of scales from returning adults indicate that most juvenile coho outmigrate as either age 1+or age 2+ but the proportion of each age c1ass has varied between years (ADF&G 1982;ADF&G 1983;Barrett et al.1984;Barrett et al.1985). Coho salmon in the middle Susitna River spawn almost exclusively in the tributaries.The fry,after emergence,re.ar in their natal tributaries or enter the mainstem river in search of suitable habitats.Outmigrant trap data collected at Talkeetna Station have shown a downstream redis- tribution of juvenile coho occurring throughout the open-water season. During the fall,coho fry move into tributaries,sloughs,beaver ponds, or other habitats to overwinter.Similar redistributions of juvenile coho were observed by Delaney and Wadman (l979)and by Tschaplinski and Hartman (l983). Trap catches recorded at Talkeetna Station during 1982 and 1984 showed that hi gh catches of age 0+and 1+juveni 1e coho occurred duri ng September or early October.It was presumed that these fish were redistributing to habitats in the lower river to overwinter,but the data collected at Flathorn Station in 1984 indicate that a portion of these fish may migrate to the ocean during the fall (Fig.22). 4.2.3 Growth The change in mean 1ength for age 0+coho by samp 1i ng peri od for the combined data collected at the Talkeetna Station outmigrant traps during 1982,1983,and 1984 is presented in Fig.52.Coho salmon in the middle river emerge from the gravel at approximately 35 mm and grow to 45 mm by early July.By the end of the open-water season,coho fry have obtained a mean length of approximately 68 rnm.Throughout the season,age 0+ coho in the lower river averaged at least five millimeters larger than fish collected in the middl~river (Fig.26). Age 1+coho sa 1mon in the mi ddl e ri ver also showed a steady growth through the season (Fig.53)increasing approximately 45 mm between late May and early October.Similar to age 0+coho,age 1+coho collected in the lower river averaged larger than fish captured in the middle river reach (Fig.27). The downstream redistribution (as shown by the cumulative biomass)of juvenile coho salmon in the Susitna River by age class during 1984 averaged one to two weeks later than the redistribution of the total number of individuals recorded at both the Talkeetna and Flathorn stations outmigrant traps (Fig.54).The difference between the cumu- lative biomass movement and the movement of total numbers of fish results from the growth of juvenile coho occurring during the open-water 83 1983 &1984 TALKEETNA COHO 1+ - 100 90 80 IJ.J 70>i= :5 60 ::l ::i ::l 50<.:l t- Zw 40Q 0:: Wa.30 20 10 0 - ..... Figure 51.Coho salmon (age 1+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. 84 - - .,~' COHO 0+MEAI'J LENGTH 70 68 66 ·64 62 (>Maxlmum...-.. :::E 60 :::.i:......,58 :::c 56I-Combined C)z 54 <>+w Minimum +...J 52 ...Jg 50 + l-48 + z 46 L5 44 (>:::.i: 42 40 38 36 34 L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP L SEP SAMPLING PERIOD OCT !~' .-. Figure 52.Coho salmon (age 0+)mean length and range of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984. " 85 60 +-----,r-------'--,-----.,----,------r---,-----r----1 L MAY E JUNE L JUNE E JULY L JULY E ,~UG L AUG E SEP L SEP 140 I ·130 J ::t:110 l- e)zw ..J 100 ..J «:( I-o I-90 z l5 2 80 70 COHO 1+M~AN LENGTH SAMPLING PERIOD j oel ~I Figure 53.Coho salmon (age 1+)mean length and range of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984. 86 - }1 )1 1 l )1 J )1 )1 j ] 1984 TALKEETNA COHO 1+ 100 90 -, "k80-l ~70 i j 60::::I :::i :J 500.- % W 400 0: Wn.30 20 10 COHO 0+1984 TALKEETNA ~o 10 50 40 90 20 70 80 30 100 I ~, ~ 3 :J::; :Jo.-zwo 0:wn. O~I ,ii,I •I l MAY E JUNE l JUNE E JULY l JULY E AUG l AUG E SEP l SEP E OCT SAMPLING PERIOD o I i I I Iii I I I l MAY E JUNE l JUNE E JULY l JULY E AUG llWG E SEP l SEP E OCT SAMPLING PERIOD ex:> -.....J 1984 FLATHORN COHO 0+ 100 90 80 w 70 2 j 60 :J::; :J 0 50 .-zw 400 0:wn.30 20 10 -l -~-----/ ~ ~ :J :::i :Jo.-z tJ 0: Wn. 1984 FLATHORN COHO 1 + 100 T-------,-------.:.--------......, 90 80 70 60 50 40 30 20 10 o I ii'I Iii I I L MA.Y E JUNE L JI)NE E JULY L JULY E AUG L AUG E SEP L SEP E OCT SAMPLlI'JG PERIOD Figure 54.Coho salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and Flathorn stations,1984. season.The cumulative biomass curve is probably a better indicator of the value of coho rearing habitat in the reach than is the cumulative numbers curve.That is,the greater the amount of time the fry spend rearing in a particular reach of river,the greater the benefit they have gained from that particular reach.Not only are they larger, having consumed more food in this reach,they also have a higher proba- bility of survival than smaller fry and therefore are of more value. Any management determination for these fish should consider the timing of movement of total biomass in the river rather than formulating actions only from the catch data. 4.3 Sockeye Salmon 4.3.1 Outmigration The migration of sockeye salmon fry past Talkeetna Station during 1984 was similar to the timing recorded during 1983 (Fig.55).Fifty percent of the total outmigration was recorded by the end of June during both seasons.Sockeye fry were steadily redistributing to areas below the sampling site from break-up through late August.Sampling of sloughs and side channels in the middle river during the cold branding study showed that sockeye fry were not actively outmigrating but were entering habitats along the margins of the river as they moved downstream.The fry probably remain at these sites until (1)they are displaced by flows or density interactions,(2)adequate food supplies are no longer available,(3)the habitats become otherwise unsuitable,or (4)the critical size is reached. The tendency of sockeye fry to ori ent along the banks of the ri ver during their downstream migration was observed at Flathorn Station where 59%of the total sockeye fry collected in the mobile trap were captured at bank transect points. The rates of downstream movement for coded wi re tagged sockeye fry during 1984 showed that fry in the middle river,after tagging,spent an average of 35 days (range from 0 to 109 days)in the middle river before migrating past Talkeetna Station. 4.3.2 Freshwater life history Outmigrant trap data collected at Talkeetna Station during the past three seasons (1982-1984)show that a 1arge number of sockeye fry migrate out of this reach as age 0+fish,but scale analysis of adult sockeye coll ected at Curry Stati on showed that thi sage cl ass repre- sented only 6.4%of the returning adults during 1984 (Barrett et al. 1985)•The 1a rgest percentage of returni ng adul ts were compri sed of fish which had spent one winter in freshwater before going to the ocean. There fore,the majority of age 0+fry from the middle river either rear in the lower river or have a low survival rate. - "'" Bernard et ale (1983)analyzed scale patterns from samples of adult sockeye sa 1mon collected from fOIJ r di fferent sites in the Sus i tna Ri ver watershed in an attempt to delineate the differences in scale patterns ..." '88 "'"' 1983 & 100 90 80 w 70 ~ ~ :3 60 =:J 2 =:J 50(,) ~zw 40<Ja::w ~30 20 10 0 1 984 TALKEETNA.SOCKEYE 0+ -- ..... Figure 55.Sockeye salmon (age 0+)adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. 89 for the period of freshwater growth for each of the sites.Samples were collected from escapements of sockeye salmon at Curry and Talkeetna stations on the Susitna River,from the outlet of Larson Lake on the Talkeetna River,and from the Tokositna River which is a tributary to the Chulitna River.One of the results of this study was that sockeye salmon scale samples collected from the Susitna River sites could not be -dtstinguished from those of Tokositna or Larson Lake fish. Six hypotheses were suggested by Bernard et ale (1983)for the lack of unique differences in the scale patterns between Susitna River fish and those collected from the other sites.In general,these hypotheses can be placed into two groups:1)The Susitna River fish are a unique stock but the fry rear in environments similar to those found in Larson Lake or the Tokositna River,or 2)the sockeye salmon spawning in the Susitna River are strays from either the Talkeetna or Chulitna watersheds and their fry move into these watersheds to rear or are displaced downstream and enter the ocean as age 0+fish.If these fish enter the ocean as age 0+fish,scale analysis of returning adults indicates that survival of these fish is very low. However,the study conducted by Bernard et ale was based on the assumption that sockeye fry did not rear in the middle Susitna River. Data collected at the Talkeetna Station outmigrant traps during the past three years have shown that a si gnifi cant amount of sockeye rea ri ng occurs in this reach.The Susitna River samples collected by Bernard et ale were taken at the fishwheel sites rather than at the spawning grounds.Barrett (1984)has pointed out that a high percentage of these fish (30%estimated in 1983)are milling fish which eventually spawned in areas other than the middle Susitna River.Comparisons of the scales of fish collected at the spawning grounds in these rivers may provide more accurate differentiation of Susitna River fish from those observed in the Talkeetna and Chulitna rivers.Also,Bernard et ale analyzed scales from only 1.3 age fish (European formula);Barrett et ale (1984) have shown that multiple age classes are present in the middle Susitna River escapements.Juvenile sockeye salmon outmigrating from Larson Lake predominantly spend two winters in freshwater before outmigrating from the lake as smolts (Mar~uson 1985). Although it is possible that sockeye salmon which spawn in the middle reach of the Susitna River are strays from the stocks originating from the Ta"lkeetna and Chulitna rivers,it is more likely that the Susitna sockeye are a separate and viable stock.However,the amount of rearing habitat in this reach is limited.The age 0+fish which outmigrate from the middle reach of the Susitna probably imprint to their natal areas in the early stages after hatching and then later distribute to suitable habitats throughout the expanse of the lower river to overwinter.These fish then enter the ocean during their second year of life and finally return to their natal areas as adults to spawn.Also,a limited amount of overwintering by sockeye fry in the middle reach does occur,as shown by the capture of age 1+fry at Talkeetna Station. More definitive information on the viability of middle Susitna River sockeye may be obtained through the continued monitoring of returning 90 - - - ,- r ! I adul ts at the fi shwheel sites and duri ng spawni ng ground surveys to collect returning fish which were marked with coded wire tags as fry. Juvenile sockeye salmon life histories in the middle Susitna River can be grouped into three categories.The first group includes those fish which spend their entire freshwater period rearing in the middle river, overwintering in this reach and then migrating to the ocean during the spring of their second year (age 1+).The second group includes those fish which rear for a portion (one to four months)of their first summer in,the middle river and then migrate to areas below the Chulitna River confluence to overwinter and then enter the ocean during the spring of their second year.The third group of juvenile sockeye spend a portion of their first surrmer rearing in the middle river and then begin a downstream mi gration,eventua 11y enteri ng the marine envi ronment duri ng their first summer or fall as age 0+fish. Currently,it is not known what contribution each group provides to the total outmigration of juvenile sockeye from the middle Susitna River. Outmigrant trap data collected at Flathorn Station during 1984 collected a large number of age 0+sockeye;most of these fish were probably destined for the ocean as 0+fish. Although trap catches of age 1+sockeye at Talkeetna Station have been low (only 19 fish during 1984),it is possible that this age class (group 1)migrates out of the middle river prior to the initiation of spring sampl ing or that they differ from their age 0+counterparts in that they migrate further from shore and are not intercepted by the bank traps in proportion to their relative abundance.Also,the bank traps are less effective at capturing these larger fish (Roth et ale 1984)~ 4.3.3 Estimate of population and survival The estimated 1983-1984 egg-to-emergent fry survival rate of 17%,based on an estimated 299,000 sockeye fry produced dur-ing 1984 from the approximately 1,900 adults which migrated past Curry Station in 1983, was lower than the 1982-1983 estimate of 42%,based on the 1,300 adult sockeye past Curry Station during 1983 which produced an estimated 575,000 fry.The substantial differences between the estimates of survival in 1983 and 1984 are due in part to the data used in the calculations.During both years,survival rates were calculated by divi ding the number of fry produced by the estimated number of eggs carried by adults past Curry Station during the previous season. Ba rrett et a 1.(1984)poi nted out that the estimates provi ded at Curry Station represent only the fish which passed this site but do not necessarily reflect the number of fish which actually spawned in the middle river reach.As sockeye salmon in this reach are almost strictly slough spawners,~ore reasonable estimates were calculated by Barrett et ale (1984)by comparing slough escapement counts to observation life data to estimate the total slough escapement in the middle river. During 1983,this comparison provided an estimate that 1,060 adult sockeye had spawned in sloughs in the middle river.The stream 1 ife data were then used to provi de comparable estimates for 1982 showi ng approximately 1,500 sockeye had spawned in the sloughs that year.These 91 data were then used to recalculate the sockeye egg-to-outmigrant sur- vival rates.A survival rate of 22%was estimated for 1983-1984 and a rate of 35%was calculated for 1982-1983. 4.3.4 Growth The weekly growth rate for sockeye fry which were coded wire tagged in 1983 and 1984 (Fig.56)most accurately represent the growth rates for sockeye salmon fry in the middle river because the dates of release and recovery and the mean lengths for each period were known. These fry grew approximately three mi 11 imeters each week unti 1 they reached a critical size and then the growth rates slowed (Fig.56). Schmi dt (1984)postul ated that the cessati on of sockeye growth after reaching a certain size was associated with evolved behavioral patterns and morphological changes.Schmidt suggested that the sockeye fry were able to rear in the middle river habitats for part of the summer but began a downstream migration in search of plankton rich environments after reaching a critical size.The small number of habitats which provide this type of environment in areas associated with the Susitna River is a major factor in controlling the production of sockeye in the middle river. A comparison of the length data collected at Talkeetna Station during 1982,1983,and 1984 and during the previous winter studies above Tal keetna in 1981 and 1982 show that Susitna River sockeye average approximately 32 mm total length at emergence,35 mm by early June,and have increased to approximately 50 mm by late July (Fig.57).From late July through August,no significant growth was observed for sockeye fry collected at Talkeetna Station,indicating that the critical size postulated by Schmidt (1984)may be 50 to 55 mm in the middle river. The apparent growth of sockeye fry after late August (Fig.57)is attributed to the collection of fish which had continued rearing in the small number of sites in the middle river which provide the necessary food and habitat requi rements.These fi sh were probably forced to migrate out of these areas as water levels and available habitat decreased.The number of sockeye collected after late August represent less than 2%of the total outmigration of age 0+fish from this reach. A comparison of the downstream redistribution of sockeye salmon in the Susitna River by age class during 1984 as the percent cumulative of the total catches recorded at Tal keetna and Fl athorn stations compared to the calculated percent cumulative biomass moving past these sites, indicated that the redistribution by weight of sockeye in the Susitna River was up to two weeks later than the redistribution observed when comparing only total numbers of fish (Fig.58). Age 1+sockeye salmon collected during 1984 averaged apprOXimately 75 mm.Thi sis apprOXimately 10 mm longer than the average 1ength of sockeye fry collected at the end of the open-water season indicating that the fry are growing through the winter and early spring prior to outmigrating as smolts.The average length of age 1+sockeye migrating out of the Susitna River was approximately 10 mm smaller than the same 92 ....1 I ~! CWT SOCKEYE MEAN LENGTH 10 56 987 7~• 85432 58 58 54-52:::E ::E 50-:J: I-48 CJ Z 48W -I -I 44 <C I-42O· I-40Z <C 38W :::E 38 34 32 1 -. ..... t"'" WEEKS BETWEEN RELEASE AND RECOVERY Figure 56.Mean length of coded wire tagged sockeye salmon fry at recovery sites in the middle reach of the Susitna River by week,1984.Number of fish shown by data points. """! 93 SOCKEYE 0+MEAN LEf'.JGTH 70,----------------------------, 65 .,.., 5 55 z ~ ....J 50g ~45 CIz L5~40 35 Maximum Q r.Io-_--1°~--____:combined+ ++ Minimum + 30 +-----,----,...----r----.,.---.....------r------.----'--l L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP L SEP SAMPlI NG PERIOD OCT Figure 57.Sockeye salmon (age 0+)mean length and range of mean lengths by sampling period recorded at the Talkeetna stationary outmigrant traps during 1982,1983,and 1984. 94 h,mll ])1 1 1 }1 t 1 J ) 1984 TALKEETNA SOCKEYE 0+1984 TALKEETNA SOCKEYE 1 + o r •I I •Ii',I l MAY E JUNE l JUNE E JULY l JULY E AUG l AUG E SEP L SEP E OCT SAMPLING PERIOD 100 90 80 ~70 ~60::I ]i ::I 500...Z hi 400 lr hi 0-30 20 10 ~.~~ 100 95 90 ~~65 ::I ::IIi ::I 600...Z hi 0 75Ii' hi Q. 70 65 60 I I •i I I Iii I L MAY E JUNE L JUNE E JULY L JULY E AUG l AUG E SEP L SEP E OCT SAMPLING PERIOD 1.0 U"l ~ ::I ::IIi ::Io... Z hio Ii' hi Q. 1984 FLATHORN SOCKEYE 0+ 100-/~.T 90 80 70 60 50 40 30 20 10 ~f i i J f I ,I j L MAY E JUNE L JUNE E JULY L ,JULY E AUG l AUG E SEP L SEP E OCT SAMPLING PERIOD ~ ~ ~ ::Io !Z ~ Ii' hi Q. 1984 FLATHORN SOCKEYE 1 + 100 /~I 90 60 70 60 50 40 30 20 10 ,I Iii J I i I I L M".Y E ,jUNE l ,JUNE (JULY L ,JULY E AUG L AUG E SEP L SEP E OCT SAMPLlI~G PERIOD Figure 58.Sockeye salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and Flathorn stations,1984. age fish outmigrating during 1984 from Larson Lake,a major spawning site in the Talkeetna River (Marcuson 1985). 4.4 Chum Salmon 4.4.1 Outmigration The migration of chum salmon fry past Talkeetna Station during 1984 was similar to the timing recorded during 1983 (Fig.59).Fifty percent of the total outmigration past this site had occurred by mid June and over 95%of the chum fry had migrated out of the middle river by mid July. At Flathorn Station,the peak chum fry outmigration also occurred in mid June during 1984. 4.4.2 Freshwater life history Chum salmon fry spend from one to eight weeks in the middle Susitna River before outmigrating from the reach.A portion of the population of chum fry probably begins outmigration shortly after emergence whereas other fry stay in the river to rear for a few weeks before outmigrating. It is not possible to determine the percentage which each group provides because of the difficulty in sampling outmigrant fishes prior to and during breakup,a time when many newly emerged chum fry may outmigrate. 4.4.3 Estimates of population and survival The estimated 1982-1984 egg-to-outmigrant fry survival rate of 16%, based on an estimated 2,039,000 chum sa 1mon fry produced duri ng 1984 from the approximately 21,100 adults past Curry Station in 1983,was similar to the estimated 1982-1983 rate of 14%,based on the 17,600 adult chum which passed Curry Station during 1982 which produced an estimated 3,322,000 fry. The calculation of survival rates is based upon the estimated number of parent spawners which is difficult to obtain because of the extent of tributary spawning by chum salmon.Also a substantial percentage of chum salmon passing Curry Station are milling fish which eventually spawn below this site,and although estimates have been provided for 1982 and 1983 (Barrett 1984),these percentages are,at best,only indicators of the amount of chum salmon milling occurring.As these estimates have a large influence on the calculated rates of survival, the rates presented for 1983 and 1984 should be used to compare differ- ences between years rather than absol ute val ues of mi ddl e river chum salmon survival. 4.4.4 Growth Many chum fry from the middle reach move downstream at lengths not much longer than their emergence length (less than 35 mm),but there are also many that spend several weeks in freshwater and attain lengths of over 60 mm,an increase of more than 20 mm.The mean 1ength by one-week periods of recovery after release for coded wire tagged chum fry which were tagged.and recaptured during 1983 and 1984 (Fig.60)most 96 - - ,~ 100 i------------------:::::::::=;;;;;=--!!!!!!!!!!::===------, ,- ~, -1 983 &1 984 TALKEETNA CHUM FRY .- ..... I 90 80 w 70>~~60 ::::J :2 ~50 I- Zw 40 <J Cl::wa.30 20 10 0-1=---1--------+----------+--------, AUG -,, - Figure 59.Chum salmon fry adjusted cumul ati ve catch recorded at the, Talkeetna stationary outmigrant traps,1983 and 1984. 97 CWT CHUM MEAN LENGTH - 5 to 10 11 to 15 1 5 to 20 2 1 to 25 DAYS BETWEEN RELEASE AND RECOVERY (Grouped by 5 Day Period) 5251~ 50 ~-:E 49:E-:::t 48 I- 0 47Z W ..J 46 ..J<45I- 0 I-44 Z<43W :E 42 41 40 o to 5 19 15 2 25 to 29 Figure 60.Mean length of coded wire tagged chum salmon fry at recovery sites in the middle reach of the Susitna River by 5 day period,1984.Number of fish shown by data points. 98 .... ! - accurately represent the growth rates of chum fry in the middle river because the dates of release and recovery and the lengths for the fish for each period were known.The 15%increase in length by fish captured more than 20 days after release (mean length significantly different from release length at 95%confidence level)would correspond to an even larger percentage increase in weight.The chum fry greater than 50 mm in length collected.during the three years of this program had a noticeably greater girth than shorter fry.Similarly,chum fry in the Tokachi River of Japan grew 1.0 to 1.3 times in length and 1.0 to 3.1 times in weight during April and May (Kaeriyama et al.1978). These data indicate that the chum fry in the middle river are actively rearing after emergence.Chum fry rearing was al so shown from the analysis of stomach samples from tagged fish recovered at Talkeetna Station during 1983.These fish had been eating various life stages of mayflies,stoneflies,blackflies,midges,and other dipterans. 4.5 Pink Salmon 4.5.1 Outmigration The rates of downstream migration of pink salmon fry past Talkeetna Station for 1983 and 1984 were very similar between the two years but the timing was approximately two weeks later in 1984 than in 1983 (Fig. 61).Differences in spawning times,winter temperatures,and spring breakup account for the differences in timing between the two years. The low catches of juvenile pink salmon recorded at Talkeetna Station during the past three seasons is due to the pattern and timing of outmigration.Pink salmon fry outmigrate shortly after emergence and most of the fry probably have migrated past the traps prior to the initiation of sampling.Those fish which are still in the middle river after breakUp appear to outmigrate in association with center channels and high velocities.I 4.5.2 Freshwater life history Pink salmon fry in the Susitna River outmigrate to the ocean shortly after emergence during a relatively short (in comparison to the other species)timing window whose boundaries are determined by the timing of spawning the previous season,incubation temperature,and the level of discharge.The pink fry collected during 1984 averaged approximately 35 111m which is similar to their mean length at emergence.A few pink fry which ranged in length from 40 to 50 mm were collected,indiCating that a small percentage of fry may be feeding for a short period of time in freshwater before outmigrating to the ocean. 99 1983 &1984 TALKEETNA PINK FRY ~, 100 _. 90 80 - w 70>i="""1 j I 60 ::J ::i ::J 50 ~,() ~zw 40() c:r::1"""1wa.30 20 10 ""'"'0 JUL AUG Figure 61.Pink salmon fry adjusted cumulative catch recorded at the Talkeetna stationary outmigrant traps,1983 and 1984. - 100 "... - - - 5.0 CONTRIBUTORS Resident and Juvenile Anadromous Fish Project Leader (Acting Project Leader, Jan.to Jun.1985) Task Leader Talkeetna Station Flathorn Station Coded Wire Tagging Cold Branding Deshka River Weir Data Base Management Jolly-Seber Model Analysis of Stomach Samples 101 Dana Schmidt Stephen Hale Kent Roth Chuck Blaney (Crew leader) Albert Badgley Patricia Harris Tom Crowe Li nda Soquet Larry Dugan (Task leader) Roger Harding (Crew leader Jeff Bigler Jim Anderson Diane Roche James Gruber Dan Sharp Linda Soquet Doug Patrick Aimee Weseman Mi ke Stratton (Task leader) Dan Gray Dave Sterritt (Crew leader) John McDonell Allen Bingham -(Project leader) Kathrin Zosel Alice Freeman Chuck Mi 11 er Donna Buchholz Stephen Hale Dana Schmidt Stephen Hale Tim Hansen (Crew leader) Craig Richards Drafting Typing Text Report Coordinators and Editors iidlll, 102 Carol R.Hepler Skeers Word Processing Kent Roth Mike Stratton Stephen Hale Drew Crawford Paul Suchanek - ~: """ - ..... ,.... - - - 6.0 ACKNOWLEDGEMENTS Funding for this study was provided by the Alaska Power Authority. We are grateful to the various consulting agencies working on the Susitna Hydroelectric Project for helpful comments on a draft of this report. 103 Adult Alaska Report .- - - - - 7.0 LITERATURE CITED Alaska Department of Fish and Game (ADF&G).1981a.Adult anadromous fisheries project (June -September 1981).Phase 1 final draft report.Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. •1981b.Juvenile anadromous fish study on the Lower Susitna----~River (November 1980 -October 1981).Phase 1 final draft report. Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. •1983a.Adult anadromous fish studies,1982.Susitna Hydro -----aquatic studies phase II final report.Volume 2.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies. Anchorage,Alaska. •1983b.Resident and juvenile anadromous fish studies on the --";:;"Susitna River below Devil Canyon,1982.Susitna Hydro aquatic studies phase II basic data report.Volume 3.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies,Anchorage,Alaska. •1985.(unpublished draft).Resident and juvenile anadromous --studies.Procedures manual draft (May 1984 -April 1985).Susitna Hydro Aquatic Studies Program,Alaska Department of Fish and Game. Anchorage,Alaska. Barrett,B.M.1984.Summary of abundance and distribution of adult salmon in Susitna river sub-bas,ins.Presented at Aquatic Habitat Workshop No.1,Susitna Hydroel ectri c Project,February 15,1984. Anchorage,Alaska, Barrett,B.M.,F.M.Thompson,and S.N.Wick (eds.).1984. anadromous fish investigations:May -October 1983. Department of Fish and Game Susitna Hydro Aquatic Studies. No.1.Anchorage,Alaska. Bernard,D.R.,G.01 iver,W.Goshert,and B.Cross.1983.Comparison of scale patterns from sockeye salmon sampled from different rivers within the Susitna River watershed in 1982.Alaska Department of Fish and Game,Division of Commercial Fisheries,Statewide Stock Biology Group.Anchorage,Alaska. Chapman,D.G.1951.Some properties of the hypergeomtric distribution with appl ications to zoological sample censuses.University of California Publication Statistics 1:131-160. Delaney,K.,K.Hepler,and K.Roth.1981.Deshka River chinook and coho salmon st~dy.Alaska Department of Fish and Game,Division of Sport Fish.Federal Aid in Fish Restoration,Project AFS-49,Vol. 22. 104 ~, Delaney,K.J.,and R.Wadman.1979.Little Susitna River juvenile chinook and coho salmon study.Alaska Department of Fish and Game, Division of Sport Fish.Anchorage,Alaska.~ Dixon,W.J.,M.B.Brown,L.Engelman,J.W.Frane,M.A.Hill,R.1. .Jennrich,and J.D.Toporek (eds.).1981.BMDP Statistical Software 1981.University of California.Berkley,California. Healy,M.e.,and W.R.Heard.1984.Inter-and intra-population variation in the fecundity of chinook salmon (Oncorh*nchus tshawytscha)and its relevance to the life history t eory. Canadian Journal of Fisheries and Aquatic Sciences 41:476-483. Kaeriyama,M.,S.Sato,and A.Kobayashi.1978.Studies on the growth and feeding habit of the chum salmon fry during seaward migration in the Tokach i Ri ver system.1.I nfl uence of thaw on the growth and feeding habit of the fry.Sci.Rep.Hokkaido Salmon Hatchery 32:27-41.(in Japanese,English summary).. Koerner,J.F.1977.The use of the coded wire tag injector .under remote field conditions.Alaska Department of Fish and Game, Informational Leaflet No.172. Kubik,S.1967.Population studies of anadromous species with emphasis on upper Cook Inlet drainage.Alaska Department of Fish and Game, Division of Sport Fish.Federal Aid in Fish Restoration,1966- 1967,Project Report 8:117-128. Kubik,S.,and K.Delaney.1980.Inventory and cataloging of sport fish waters of the lower Sus·itna River and central Cook Inlet drainages.Alaska Department of Fish and Game.Federal Aid in Fi sh Restorat ion,Annual Report of Progress,1979-1980,Proj ect F-9-12,21(G-I-H). Kubik,S.,and R.D.Wadman.1978.Inventory and cataloging of sport fish waters of the lower Susitna River and central Cook Inlet drainages.Alaska Department of Fish and Game.Federal Aid in Fish Restoration,Annual Report of Progress,1978-1979,Project F-9-11,20(G-I-H). Manly,B.F.J.1984.Obtaining confidence limits on parameters of the Jolly-Seber model for capture-recapture data.Biometrics 40:749-758. Marcuson,P.1985.Larson Lake project progress report.Cook Inlet Aquaculture Association,Anchorage,Alaska. McConnell,R.J.,and G.R.Snyder.1972.Key to field identification of anadromous juvenile salmonids in the Pacific Northwest.National Oceanic and Atmospheric Administration Technical Report,National Marine Fisheries Service CIRC-366. 105 - ~, - - - - - .-I Moberly,S.A.,R.Miller,K.Crandall,and S.Bates.1977.Mark-tag manual for salmon.Alaska Department of Fish and Game.Fisheries Rehabilitation and Enhancement Division. Morrow,J.E.1980.The freshwater fishes of Alaska.Alaska Northwest Publishing Company,Anchorage,Alaska. Raleigh,R.F.,J.B.McLaren,and D.R.Graff.1973.Effects of topical location,branding techniques and changes in hue on recognition of cold brands in Centrarchid and Salmonid fish.Transactions of the American Fisheries.Society 102:637-641. Richards,K.1979.Aspects of the juvenile life history of spring chinook salmon (Oncorhynchus tshawytscha)in Deshka River,Alaska determined from adult scale analysis and migrant trapping.M.S. Thesis.Oregon State University,Corvallis,Oregon. Ricker,W.E.1975.Computation and interpretation of biological statistics of fish populations.Bulletin of the Fisheries Research Board of Canada.191. Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of juvenile salmon from the Susitna River above the Chulitna River confluence.Part loin D.C.Schmidt,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds--:).1984.Resident and juvenile anadromous fish investigations (May -October 1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska Department of Fish and Game. Anchorage,Alaska. Schaefer,M.B.1951.Estimation of the size of animal populations by marking exper.iments.United States Fish and Wildlife Service, Fisheries Bulletin 52:189-203. Schmidt,D.C.1984.Riverine rearing of slough spawned sockeye salmon in the Susitna River.Paper presented at annual meeting of the American Fisheries Society,Alaska Chapter.November,1984. Juneau,Alaska. Schmidt,D.C.,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.). 1984.Resident and juvenile anadromous fish investigations (May - October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage,Alaska. Trautman,M.B.1973.A guide to the collection and identification of pre-smolt Pacific salmon in Alaska with an illustrated key. National Oceanographic and Atmospheric Administration Technical Memorandum.NMFS ABFL-2. Tschaplinski,P.J.,and G.F.Hartman.1983.Winter distribution of juvenile coho salmon (Oncorhynchus kisutch)before and after logging in Carnation Creek,British Columbia,and some implications for overwinter survival.Canadian Journal of Fisheries and Aquatic Sciences 40:452-461. 106 Vining,L.J.,J.So Blakely,and GoM.Freeman.198.50 An evaluation of the incubation 1 ife-phase of chum salmon in the middle Susitna River,Alaska.Winter Aquatic Investigations:September,1983 - May 1984.Susitna Hydro Aquatic Studies.Report No.50 Alaska Department of Fish and Game,Anchorage,Alaska. 107 - - - - - APPENDIX A JUVENILE SALMON CATCH AND LENGTH DATA,1984 Appendix Table A-1.Weir catches of juvenile chinook and coho salmon on the Deshka River, May 10 through September 19,1984.- Chinook Coho Tributary Hours Daily Catch Dai ly Catch Date River Mile Fished Catch Per Hour Catch Per Hour ,-May 10 2.0 21.5 2 0.1 0 0.0 12 2.0 15.0 9 0.6 1 0.1 13 2.0 21.0 3 0.1 0 0.0 27 5.0 12.0 50 4.2 1 0.1 28 5.0 12.5 7 0.6 0 0.0 r-29 4.5 12.5 3 0.2 0 0.0 31 5.0 12.0 4 0.3 0 0.0 June 1 5.0 12.5 21 1.7 0 0.0 ~.21 5.0 11.5 1 0.1 0 0.0 22 5.0 21.5 3 0.1 0 0.0 July 11 2.5 14.5 209 14.4 5 0.3 12 2.5 24.0 144 6.0 2 0.1 13 2.5 24.0 268 11.2 3 0.1 14 2.5 23.5 186 7.9 4 0.2 15 2.5 24.0 27 1 .1 0 0.0 16 2.5 24.0 130 5.4 1 0.0 ~25 2.5 15.0 318 21.2 21 1.4 26 2.5 24.0 149 6.2 8 0.3 31 2.5 20.0 168 8.4 4 0.2 August 13 2.5 14.0 45 3.2 15 1 .1 14 2.5 23.0 4 0.2 2 0.1 15 2.5 23.0 5 0.2 5 0.2 16 2.5 23.0 27 1.2 12 0.5 31 2.0 21.5 5 0.2 22 1.0 September 11 1.5 13.5 1 0.1 0 0.0 12 1.5 23.0 6 0.3 0 0.0 13 1.5 23.0 8 0.3 1 0.0 14 1.5 23.0 2 0.1 0 0.0 15 2.5 18.0 1 0.1 2 0.1 16 2.5 24.0 0 0.0 6 0.3 17 2.5 24.0 1 0.0 0 0.0 18 2.5 23.0 1 0.0 2 0.1- Season Totals 621.0 1,808 2.9 117 0.2 ,- A-I Appendix Table A-2.Results of incidental minnow trapping in the Deshka River.1984. - Chinook Coho Tributary Number Catch Catch River Hours of Daily Per Daily Per Date Mile Fished Traps Catch Trap Catch Trap June 21 5.5 16 6 56 9.3 14 2.3 ..... August 28 2.5 9 o 6 15 2.5 48 8.0 29 2.7 7 7 23 3.3 50 7.1 September 17 5.5 24 4 20 5.0 4 1.0 -, October 10 2.2 24 2 1 0.5 2 1.0 10 6.0 24 4 30 7.5 4 1.0 11 5.0 27 7 23 3.3 21 3.0 13 2.0 to 6.0 54 5 2 0.4 10 2.0 14 2.0 to 6.0 28 5 1 0.2 4 0.8 15 4.0 24 5 41 8.2 9 1.8 ~I Season Totals 51 212 4.2 166 3.3 -, A-2 ))}1 1 --,1 1 ~-~1 1 J 1 Appendix Table A-3.Mean length and range of lengths for age 0+chinook salmon by sampling period in the lower reach of the Susitna River,1984. Lower Susitna Flathorn Station Deshka River JAHS Sitesa Sampling Period Mean Range of Mean Range of Mean Range of n Length Lengths n Length Lengths n Length Lengths May 0 --77 42.7 36-49 b June 1-15 24 56.6 40-67 21 42.4 40-46 74 48.5 34-63 June 16-30 374 58.5 39-74 56 55.7 46-69 63 52.0 36-70 July 1-15 357 62.0 40-84 236 66.8 52-83 84 54.5 39-74 July 16-31 436 64.3 43-88 201 69.7 52-93 171 58.1 39-80 August 1-15 189 66.6 47-89 53 74.4 60-91 330 58.9 40-82 »August 16-31 193 72.7 46-94 65 71.7 55-89 238 61.5 42-94Iw September 1-15 8 77.3 68-84 15 77.9 69-88 52 66.8 52-95 September 16 -October 15 10 78.7 68-95 102 76.0 68-85 53 73.2 51-92 -a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the Chulitna River confluence. b Not sampled. Appendix Table A-4.Mean length and range of lengths for age 0+chinook salmon by sampling period in the Talkeetna River and the middle reach of the Susitna River,1984. Talkeetna River Talkeetna Station Middle Susitng Indian RiverMarkingSites Sampling Period Mean Range of Mean Range of Mean Range of Mean Range of n Length Lengths n Length Lenghts n Length Lengths Ii Length Lengths May b --2 55.5 53-58 60 40.8 35-45 b June 1-15 0 --54 48.6 36-66 b - - b June 16-30 26 52.2 43-64 475 53.0 37-70 b --b July 1-15 159"56.0 44-70 538 56.2 38-75 100 47.8 38-67 50 48.9 42-64 Jul Y 16-31 155 56.1 40-74 1131 55.5 37-80 50 52.2 42-69 50 54.9 47-67 August 1-15 257 60.7 44-84 748 57.9 40-90 50 52.4 40-77 100 58.8 47-90 >August 16-31 114 65.2 51-84 612 59.5 39-95 100 56.1 43-72 100 61.1 49-80I.,t::. September 1-15 0 --119 62.7 45-91 100 57.6 47-88 100 63.8 47-90 September 16 -October 15 b - - 13 60.8 51-90 200 61.0 45-90 300 65.5 50-89 - a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna River between the Chulitna River confluence and Devil Canyon. b Not sampled. ,J )J )1 J »)J J 1 ),)_J 1 ,)J ))1 1 1 "1 J -)..-1 -J l ).,1 1 Appendix Table A-5.Mean length and range of lengths for age 1+chinook salmon by sampling period in the Susitna River,1984. Flathorn &Talkeetna Flathorn Station Talkeetna Stations Stations Combined Sampl i ng Mean Range of Mean Range of Mean Range of Period n Length Lengths n Length Lengths n Length Lengths May 11 79.7 67-105 209 77.9 61-101 220 78.0 61-105 Early June 104 89.1 70-122 126 89.6 71-112 230 89.7 70-122 Late June 101 85.2 75-122 335 88.4 71-107 436 87.7 71-122 Early July 17 94.1 86-113 218 85.7 76-117 235 86.3 76-117 Late July 4 97.5 95-102 96 87.7 81-115 100 88.1 81-115 Early August 8 98.6 90-113 1 91.0 91 9 97.8 90-113 Late August 2 96.0 95-97 0 - - 2 96.0 95-97 ~ I U1 Appendix Table A-6.Mean length and range of lengths for age 0+coho salmon by sampling period in the lower reach of the Susitna River,1984. Flathorn Station Sampling Period Mean Range of n Length Lengths n May 0 --0 June 1-15 10 42.7 32-60 0 June 16-30 19 48.7 32-64 0 July 1-15 11 49.3 36-65 0 July 16-31 38 58.6 44-73 21 August 1-15 30 62.1 49-79 19 August 16-31 181 66.8 40-89 59 :i- &September 1-15 84 75.0 55-94 2 September 16 -October 15 67 75.1 57-94 29 Deshka River Mean Range Of Length Lengths 57.3 47-65 63.6 53-72 71.2 51-89 68.0 67-69 77.0 60-95 Lower Susitna JAHS Sitesa Mean Range of n Length Lengths b 18 40.9 33-50 9 46.2 34-61 26 50.7 35-65 33 50.2 37-65 45 49.6 41-68 71 59.1 40-85 59 62.2 49-86 105 66.7 49-95 a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the Chulitna River confluence. b Not sampled. I J •,J .J J J .J J )t I "J ~J t , 1 1 "1 ])]J -1 --1 )1 1 ] Appendix Table A-7.Mean lengths,and range of lengths for age 0+coho salmon by sampling period in the middle reach of the Susitna River,1984. Talkeetna Station Sampling Period Mean Range of n Length Lengths May 35 39.7 35-46 June 1-15 40 39.6 30-51 June 16-30 156 43.9 31-58 July 1-15 242 47.8 32-63 July 16-31 439 51.8 33-69 August 1-15 221 54.1 41-74 )::-August 16-31 198 61.5 42-80 I '-l September 1-15 212 60.5 42-85 September 16 -October 15 39 69.1 51-90 Middle Susitn9 Indian RiverMarkingSites Mean Range of Mean Range of n Length Lengths n Length Lengths b --b b --b b --b 0 - - 62 38.0 34-51 0 - - 10 44.1 42-49 0 --80 48.0 39-58 38 50.8 39-62 46 49.0 42-61 41 56.8 40-70 90 50.9 44-64 5 59.4 48-76 166 55.1 44-73 a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna River between the Chulitna River confluence and Devil Canyon. b Not sampled. Appendix Table A-8.Mean length and range of lengths for age 1+coho salmon by sampling period in the lower reach of the Susitna River,1984. Lower Susitna Flathorn Station Deshka River JAHS Sitesa Sampling Period Mean Range of Mean Range of Mean Range of n Length Lengths n Length Lengths n Length Lengths May 0 --5 69.8 58-89 b June 1-15 7 87.4 62-110 0 --1 70 70 June 16-30 15 78.1 65-96 14 78.6 58-108 11 97.4 62-111 July 1-15 12 84.9 70-111 13 79.0 62-95 6 81.3 72-101 July 16-31 39 89.8 75-120 6 101.7 65-118 4 85.3 73-92 August 1-15 16 92.8 80-112 2 97.5 83-112 4 102.0 98-109 August 16-31 68 103.4 91-122 68 98.2 90-123 11 105.2 90-123 ):0 I September 1-15 68 109.4 95-129 1 118.0 118 3 105.3 104-108co September 16 - October 15 53 112.9 95-133 31 111.8 92-134 4 112.0 99-110 a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the Chulitna River confluence. b Not sampled. ~-,I J J ),,~J I J J ~~.t J )~.) J 1 J 1 J 1 )J 1 1 )1 I Appendix Table A-9.Mean lengths,and range of lengths for age 1+coho salmon by sampling period in the middle reach of the Susitna River,1984. Middle Susitng Marking Sites Mean Range of length Lengths b b b 2 67.0 64-70 7 85.7 79-90 17 86.1 74-99 0 0 0 Talkeetna Station Sampling Period Mean Range of n length lengths May 139 69.4 51-105 June 1-15 332 71.8 52-102 June 16-30 340 76.1 59-115 July 1-15 192 77.8 64-118 July 16-31 252 82.2 70-125 August 1-15 28 93.5 79-120 ):=0 August 16-31 96 101.9 81-131 I 1.0 September 1-15 14 99.6 86-127 September 16 -October 15 21 114.4 93-135 n 18 b b o o o 2 10 4 63.0 103.5 93.2 93.5 52-85 102-105 83-101 90-99 n Indian River Mean Range of Length lengths a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna River between the Chulitna River confluence and Devil Canyon. b Not sampled. Appendix Table A-10.Mean length and range of lengths for age 2+coho salmon by sampling period on the Susitna River between Cook Inlet and Devil Canyon,1984.~ ~~ Sampling n Mean Range of Period Length Lengths -. May 5 133.2 120 -160 ~ E.June 7 135.6 114 -157 L.June 1 136.0 136 ~ E.July 2 130.0 130 L.July 0 E.August 1 126.0 126 ""'"L.August 13 138.0 125 -176 E.September 2 134.0 134 ""'" L.September - E.October 13 141.0 135 -150 ~ All Season 44 137.1 114 -176 ~ A-IO ~, Appendix Table A-11.Daily catches of outmigrant chum and sockeye salmon fry in a fyke net located at the mouth of Slough 21,May 23 to June 12,1984. Check Date Sockeye Chum Check Date Sockeye Chum May 23 1,005 74 June 3 155 8 24 694 83 4 140 8,- 25 810 60 5 164 10 26 2,150 355 6 419 12 27 1,479 399 7 1,024 82 28 400 83 8 570 85 29 1,777 198 9 761 59 30 253 89 10 31 34 ~.June 156 44 11 23 8 2 344 33 12 29 8 13a 2 1 a Slough breached allowing fish passage around net.Net pulled. r .... A-ll Appendix Table A-12.Mean length and range of lengths for age 0+sockeye salmon by sampling period on the Susitna River between Cook Inlet and Devil Canyon,1984. Lower Susitna a Middle Susitna b Flathorn Station JAHS Sites Talkeetna Station Marking Sites Sampling --Mean MeanPeriodMeanRangeofMeanRangeofRangeof Range of n Length Lengths n Length Lengths n Length Lengths n Length Lengths May 134 32.8 27-45 c --213 32.0 26-41 100 30.5 25-37 June 1-15 284 40.4 29-60 15 36.0 26-52 305 36.5 28-60 100 35.2 29-49 June 16-30 343 42.7 25-70 80 40.1 26-66 509 41.9 25-71 50 34.2 28-44 JUly 1-15 313 49.2 25-8p 20 43.6 30-65 570 48.8 30-75 0 July 16-31 337 52.2 30-85 54 43.5 28-76 748 53.4 35-87 8 53.1 47-68 August 1-15 239 53.0 29-85 38 47.9 30-76 547 51.8 33-88 49 51.4 43-62 ):-August 16-31 185 52.8 30-93 106 53.0 28-86 90 58.6 42-79 50 56.2 36-69 I...... N September 1-15 41 55.6 42-75 20 61.2 45-71 95 59.8 40-91 0 September 16 -October 15 37 57.2 38-81 62 60.3 35-79 15 60.4 48-90 0 a Includes all mainstem,slough,and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the Chulitna River confluence. b Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna River between the Chulitna River confluence and Devil Canyon. c Not sampled. )I _J __J _c_J ~J J '. )J •J -)J J J l /'~ I~ , Appendix Table A-13.Mean length and range of lengths for age 1+ sockeye salmon by sampl ing period on the Susitna River between Cook Inlet and Devil Canyon,1984. A-13 Appendix Table A-14.Mean length and range of lengths for chum salmon fry by sampling period on the Susitna River between Cook Inlet and Devil Canyon,1984. Lower Susitnaa Middle Susitna b F1athorn Station JAHS Sites Talkeetna Station Marking Sites Sampling Range of Mean Range of Range ofPeriodMeanRangeofMeanMean n Length Lengths n Length Lengths n Length Lengths n Length Lengths May 35 42.7 36-62 c --367 40.1 32-52 150 39.9 33-47 June 1-15 198 41.9 30-55 298 43.2 31-58 357 45.6 35-68 300 44.5 36-60 June 16-30 209 42.7 32-63 109 39.4 31-50 427 42.9 36-62 50 40.2 36-48 July 1-15 17 42.5 30-59 37 42.3 33-57 337 44.0 35-65 50 48.2 39-54 July 16-31 3 43.3 31-52 21 40.4 36-47 172 44.6 36-59 10 46.5 40-51 Includes all mainstem,slough,and side channel sites sampled during the JAHS studies in the Susitna Riyer between Cook Inlet and the Chulitna River confluence. a b Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna River between the Chulitna River confluence and Devil Canyon. c Not sampled. » I..... ~ ~J J J 1 ,t I !J .-j J ~jJ ~j ~)j - APPENDIX B THE SCHAEFER ESTIMATE OF POPULATION SIZE The Schaefer method of estimating population size is useful with migrating fish which can be sampled and marked at one point and recovered later at a different point on the migratory route (Ricker 1975).The Schaefer estimate of population size (N)is given by Ricker as: number of fish marked during a single tagging period. total recaptures of fish tagged in the ith period number of fish captured and examined for marks during a recovery period. N =2'N..=~0".Mi .Cj )lJ L..lJ r If:" 1 J number of fish which were marked during a tagging peri od (i)and subsequently recaptured duri ng a recovery period (j). where:R..=lJ F"'. Mi = R.= 1 C.= J R.=number of marked fish which were recaptured during a J recovery period. Nij =estimate of the number of fish available for marking during a period (i)and the number available for recovery in a period (j). ..- - Tagging and recovery periods for the Susitna River study were grouped by eight-day intervals.The data collected for the estimate of the popu- lation of sockeye salmon outmigrants are tabulated by the Schaefer method in Appendix Table B-1.The computation of the population esti- mate is presented in Appendix Table B-2 . Because only age 0+sockeye fry were tagged and because some of these remained in the middle river to overwinter (therefore,there was no chance of recapturing them as age 0+fry at Tal keetna Station),we had to assume that the marked/unmarked ratio was the same for the fry that outmigrated as it was for the fry that remained to overwinter.The purpose of sampl ing at Talkeetna Station was to estimate this ratio. Data collected so far indicate that the number of overwintering sockeye fry in this reach is low in comparison to the number that outmigrate,so the consequences of violating this assumption are not severe. The mark-recovery data for chum salmon are presented in Appendix Table B-3,and the computations and final population estimate are provided in Appendix Table B-4. - B-1 Appendix Table B~1.Data collected on the coded wire tag,mark~recapture experiment for sockeye salmon fry to provide a Schaefer population estimate.Tagging and recovery periods are by eight day intervals,May 22 through September 18, 1984. Period of Period of Tagging (i)Tagged Fish Total Fish Recovery Recovered Recovered (j)1 2 3 4 (Rj)(Cj)Cj/Rj 1 27 ~--27 339 12.6 2 4 --~4 71 17.8 3 7 -~-7 414 59.1 4 26 -6 5 37 1,293 34.9 5 21 -5 24 50 931 18.6 6 70 -16 15 101 1,627 16.1 7 32 -9 7 48 976 20.3 8 16 .1 3 20 428 21.4 9 29 -5 10 44 693 15.8 OJ 10 6 2 4 12 360 30.0I- N 11 6 -~-7 173 24.7 12 - - 1 -1 20 20.0 13 1 ---1 46 46.0 14 2 ~--2 60 30.0 15 1 ---1 31 31.0 Total Tagged Fish Recovered (Ri)248 0 45 69 362 7,462 Total Fish Tagged (Mi)8,795 0 2,052 3,685 14,532 Mi /Ri 35.5 -45.6 53.4 ~_~J J J 1 I J 1 o_J !),j )J J 1 J ..- Appendix Table B-2.Computation of the sockeye salmon for outmigrant population from the data presented in Appendix Table B-1 • B-3 Appendix Table B-3.Data collected on the coded wire tag,mark-recapture experiment for chum salmon fry to provide a Schaefer population estimate.Tagging and recovery periods are by eight day intervals,May 22 through July 24,1984. Period of Period of Tagging (i)Tagged Fish Total Fish Recovery Recovered Recovered (j)1 2 3 4 (Rj)(Cj)Cj/Rj 1 11 ---11 932 84.7 2 -1 --1 104 104.0 3 3 4 2 -9 860 95.6 4 -3 3 6 12 526 43.8 5 1 3 -8 12 361 30.1 6 ---1 1 334 334.9 7 ---4 4 154 38.5 8 - - - 1 1 132 132.0 OJ Total Tagged I Fish Recovered ~(Ri)15 11 5 20 51 Total Fish Tagged (Mi)4,806 12,276 5,295 9,019 31,396 Mi/Ri 320.4 1 )116.0 1,059.0 451.0 )I .~)J )J J J ~,.1 ,J • Appendix Table B-4.Computation of the chum salmon for outmigrant population from the data presented in Appendix Table B-3. Period of Tagging (i) -.. ".... .... - Period of Recovery (j) 1 2 3 4 5 6 7 8 TOTAL 298,517 91,891 9,644 400,052 2 116,Q64 426,758 146,642 100,775 790,239 B-5 3 202,481 139,153 341,634 4 118,523 108,601 150,634 69,454 59,532 506,744 Total 298,517 116,064 721,130 404,318 219,020 150,634 69,454 59,532 2,038,669 ~, .- APPENDIX C Time Series Analysis of Discharge, Turbidity,and Juvenile Salmon Outmigration in the Susitna River,Alaska - - - TIME SERIES ANALYSIS OF DISCHARGE,TURBIDITY,AND JUVENILE SALMON OUTMIGRATION IN THE SOSITNA RIVER,ALASKA by:Stephen S.Hale Alaska Department of Fish and Game Susitna River Aquatic Studies Program 620 East lOth Avenue,Suite 302 Anchorage,Alaska 99501 ABSTRACT During the three years of study of juvenile salmon outmigration from the middle reach of the Susitna River,a correspondence has been noted between the peaks of river discharge and the peaks of outmigration. Further investigation of the relationship of outmigration to discharge was required because two large hydroelectric dams have been proposed for a region above the salmon rearing areas.These dams will markedly change the downstream discharge and turbidity regimes,factors which influence not only salmon outmigration, but almost all fish species and life stages including juvenile salmon rearing.Box-Jenkins models were developed for the 1983 and 1984 time series of river discharge,tur- bidity,and chinook and sockeye salmon fry outmigration rates in order to better understand the forces that shape the seri es and to stati s- tically describe the natural conditions as a baseline against which future changes can ,be measured.The time series examined were described by relatively simple models,using mostly first-order autoregressive terms.About 85%of the variance in turbidity for one day was explained by the value for turbidity of the previous day.This figure was 44%for chinook salmon outmigration and 43%for sockeye salmon outmigration,the lower numbers indicating the effect of behavioral decisions on bio- logical time series.Although the form of the time series plots of discharge and chinook salmon outmigration was different between the two years,the underlying stochastic processes which generated these series were the same.Bivariate transfer function model s were constructed for turbidity and salmon outmigration rates which explain present values of these variables in terms of their own past values as well as past values of discharge. - - .-, -~ TABLE OF CONTENTS ABSTRACT ••.•e·'.• • • • • • • • • •i LIST OF FIGURES ••••••••••••••••••••••••••••••••••••••••••••••.•••••i1-i 1.0 INTRODUCTION •••••••••••••••••••••••••••••••••••_...............1 1.1 Time Series Analysis •••.•••.••••••.•••••••••~................4 1.2 Applications of Time Series Analysis........................5 1.3 O-bjectives,•••'•.,••••••••••-•••,•••••••••••••••••••••••••••••••.,5 2.0 METHODS •••••••••••••••••••••••••••••__•••• ••••••••••••••••••••• _7 2.1 The Data ••••••••••••••••••••••••••••••••••••••••••.••••••'...7 2.2 Identification and Estimation of Time Series Models.........7 2.3 Transfer Function Models....................................10 3.0 RESUL 1S •••••••'•••••••••••••••••••••••••••••••••• •• ••••••••••••11 3.1 Univariate Model for Mean Daily Discharge...••••••••••••••••11 3.2 Univariate Model for Turbidity..............................17 3.3 Univariate Model for Age 0+Chinook Salmon Outmigration.....22 3.4 Univariate Model for Age 0+Sockeye Salmon Outmigration.....31 3.5 Discharge -Turbidity Transfer Function ModeL..............31 3.6 Discharge -Chinook Transfer Function Model ••••~............35 3.7 Discharge -Sockeye Transfer Function ModeL................37 4.0 DISCUSSION.....................................................40 5.0 ACKNOWlEDGEME-NTS •••••••••••••••••••••••••••••••••••••••8.• • • • •44 6~O LITE.RATURE CITED..............................................45 7.0 BOX-JENKINS ARIMA AND TRANSFER FUNCTION MODELS................50 ii - ----~-~---------_._-- - LIST OF FIGURES Figure Title 1 2 3 Map of the Susitna basin study region ..•.•..•••••.•....•• Discharge,turbidity,and chinook and sockeye salmon outmigration rate,1983 ••••••••••.•••••••••••.••.. Discharge,turbidity,and chinook and sockeye salmon outmigration rate,1984 .••••••••.•.•••.••••••.•••• 2 8 9 - Susitna River discharge time series at the Gold Creek gaging station,1983 and 1984......................12 5 4 Plots of autocorrelations and partial auto- correlations for 1983 discharge time series..............13 6 Log-transformed discharge time series,1983 and 1984 'COl •••••••••••••••••••0 •e _II Ql 14 7 Plots of autocorrelations and partial auto- correlations for 1983 log-transformed discharge time series .•.•.•...•.•........•.•.•......•.•......eoGo.o.15 8 Spectrum of 1983 discharge time series...................16 9 10 Plots of autocorrelations and partial auto- correlations for 1984 discharge time series..............18 Plots of autocorrelations and partial auto- c~rrelat!ons for 1984 log-transformed discharge tlme serles B •••••••••••••••••a..................19 ~, 11 Turbidity time series at Talkeetna Station, 1983 and 1984 o..os.................20 12 Plots of autocorrelations and partial auto- correlations for 1983 turbidity time series..............21 13 Differenced turbidity time series,1983..................23 14 Plots of autocorrelations and partial auto- c~rrelat!ons for differenced 1983 turbidity tlme ser1es..............................................24 - 15 Age 0+chinook salmon outmigration rate time series,1983 and 1984 ..e.................................25 16 Plots of autocorrelations and partial auto- correlations for 1983 chinook salmon outmi- gration time serles......................................26 ..... iii --. LIST OF FIGURES (Continued) Figure 17 18 19 20 Title Page Log-transformed age 0+chi nook salmon outmi -. gration rate,1983 and 1984..............................28 Plots of autocorrelations and partial auto- correlations for log-transformed 1983 chinook salmon outmigration time serles..........................29 Plots of autocorrelations and partial auto- correlations for log-transformed 1984 chinook salmon outmigration time series..........................30 Age 0+sockeye salmon outmigration rate time series,1983 and 1984....................................32 22 21 Plots of autocorrelations and partial auto- correlations for 1984 sockeye salmon outmi- gratian time series......................................33 Plot of cross correlations between the resi- duals of the ARMA (1,1)discharge model and the prewhitened turbidity time series,1983 data.............34 23 Plot of cross correlations between the residu- als of the ARMA (1,1)discharge model and the prewhitened chinook salmon outmigration time series,1983 data .......•.........-.......................36 - .- 24 Plot of cross correlations between the residu- a 1s of the ARMA (1,1)di scha rge model and the prewhitened sockeye salmon outmigration time series,1984 data ••••••••••••,•••••••••••••~..............38 iv - - - 1.0 INTRODUCTION While examining the plots of daily catch rate of outmigrating juvenile salmon at the Talkeetna Station outnligrant traps,an apparent correspon- dence was noted between the peaks of the time series of mean daily discharge and the time series of salmon outmigration (Hale 1983;Roth et al.1984).Correlation analysis showed that there was a relatively strong relationship between discharge and the outmigration rates of various species/age classes of salmon during certain periods of time. The term outmi grati on rate is used here to mean the number of outmi- grating fry captured at the traps per hour,not the distance travelled per hour.This relationship is not simply a matter of a greater volume of water being fished at higher discharges.The correlations of catch rate of age 0+salmon with water velocity at the mouths of the traps were not significantly different from zero (Roth et al.1984,Appendix A).There was in fact a greater number of fry per unit volume of water at high levels of discharge than at low levels. A correspondence between discharge rate and salmonid outmigration has also been reported by other investigators (Cederholm and Scarlett 1982 - coho salmon;Congleton et al.1982 -chum and chinook salmon;Godin 1982;Grau 1982;Solomon 1982b).The selective advantages of this behavi.or,according to Solomon (l982b),include easier passage over long distances or shallow areas and protection from predators provided by increased turbidity and by the large numbers resulting from a coor- dinated mass migration in response to an environmental cue. There are probably two mechanisms which account for this relationsh"jp in the Susitna River.One is that the fish,which have gradually become physiol ogica lly ready for outmi grati on by growth and in response to photoperiod and temperature,are stimulated by a rise in mainstem discharge to begin that outmigration (Grau 1982).The second mechanism is that high flows physically displace the fish downstream.This latter mechani sm may frequently occur for fry rea ri ng ins i de slough s,pa rt icu- larly for chum salmon (Oncorhynchus ketal and sockeye salmon (0.nerka). The natal sloughs for many chum anCfSOckeye sa.lmon have berms at the heads which prevent water from the mainstem from entering the site at low levels of discharge.When high flows occur,the slough heads are overtopped and the fry which had been rearing in low velocity water are subjected to a strong current. Because two large hydroelectric dams have been proposed for the Susitna River in an area upstream of the rearing areas of the juvenile salmon (Fig.1),and because these dams would markedly alter the natural dis- charge and turbidity regimes,it is necessary to quantify the relation- ship between the di scharge and turbidity regimes and the outmi gration patterns of the juvenile salmon.After the dams begin operation,the annual patterns of river discharge and turbidity level would be smoothed -both would be lower than normal in the SUmnler and higher than normal in the winter.Also,the high frequency (daily)oscillations of these two time series would be dampened;there would be less day to day variation. 1 N INLET f7 •10 RIIIe,mll,Incn""nis •P,opoilid Dam Figure 1.Map of the Susitna basin study region. Data Center). (Source:Arctic Environmental Information J I J I ".,.J 5 I _I j )~J j )J )J -I - - -- There are many factors other than di scharge and turbi di ty whi ch affect the outmigration timing of juvenile salmon including time of year,size of fish,photoperiod,light intensity,and temperature (Brannon and Sa10 1982);however,discharge and turbidity bear further investigation because of the changes in these two va ri ab 1es whi chwou1 d be caused by the proposed dams.Changes in river flow can affect the survival rate of young salmon (Stevens and Miller 1983).Potential negative effects of an altered flow regime include accelerated or delayed timing of outmigrations.Changes in outmigration timing may place the fish in their rearing areas at an unfavorable time from the standpoint of food supply,which could cause reduced survival (Hartman et a1.1967).Lower discharge levels can result in a shorter distance covered per day (Raymond 1968).Decreasing mainstem flows can lead to stranding of fish in pools which have been isolated from the mainstem (Solomon 1982a). Lower flows and clearer water than normal may also result in increased predation (Stevens and Miller 1983). Turbidity level in the Susitna River probably does not have much direct effect on the daily number of fry which outmigrate or on the initiation of outmigration.In clear water streams,however,an increase in turbidity level can directly increase the number of outmigrating salmon by providing cover from predators (Solomon 1982b).Turbidity level in the Susitna River does change outmigration timing because fry in turbid water outmigrate during the day as well as during the night (Godin 1982; Roth et a1.1984).Clearing of the water.cou1d force the fry to shift to a nocturnal outmigration to avoid predators.However,this would be of marginal benefit for fry during the continuous daylight in June and July at 63 0 N latitude. To avoid or alleviate the above problems,it is necessary to understand the mechanisms producing the present discharge,turbidity,and outmi- gration regimes.Knowledge of the discharge-outmigration relationships will be useful in trying to establish a post-project flow regime which will not interfere with the natural outmigration timing. Also,because discharge and turbidity level are important variables affecting salmon life stages other than the outmigration phase as well as other species,it is necessary to statistically describe the natural discharge and turbidity regimes as a baseline against which .future changes in these variables can be measured.Turbidity provides cover for salmon fry (Suchanek et a1.1984;Part 2 of this report)but also decreases primary producti on and affects the feeding,movement,and distribution of many of the fish species present in the river.Turbi- dity level after the dams begin operation will not only be influenced by a changed discharge regime,but will also be directly changed by the dams because settling of suspended sediment in the reservoir will create a turbidity regime substantially different from the present regime. Turbidity was included as a variable of interest in this paper more because of its effect on other life stages and species than because of its effect on salmon outmigration. Further,discharge is the major variable in the extensive instream flow habitat modeling effort which has been conducted in the Susitna River; turbidity is also an important factor (Hale et a1.1984;Suchanek et ale 3 1984;Part 2 of this report).The current discharge and turbidity regimes that are driving these habitat models must be accurately described so that the models can be put into a proper perspective. 1.1 Time Series Analysis The statistical methods collectively known as time series analysis.are a logical choice for analyzing the natural discharge,turbidity,and outmigration regimes.A time series is a collection of observations ordered in time such as daily water temperature measurements.Time series analysis includes frequency domain (spectral analysis)and time domain problems.Spectral analysis is concerned with transforming a time series with a Fourier transform to a sum of sines and cosines (see Priestley 1981)and is appropriate with periodic·series such as the classical example of the Canada lynx/snowshoe hare ten year cycle (Bulmer 1978).Methods for time domain problems (or Box-Jenkins models) a re referred to as ARIMA (autoregressive,integrated,moving average) models (Box and Jenkins 1976).ARIMA models have been used extensively in economic forecasting (Nelson 1973;Granger and Newbold 1977). Time series are shaped by both deterministic and stochastic (random) events.The series has a II memo ry II of the random events (or Il s hocks ll ) operating on the series,that is,the effect of these disturbances may be apparent for several time units after the event occurred.One aspect of time series analysis consists of removing deterministic trends from a time series so that the values fluctuate around a mean level.A trans- formation may be necessary to ensure a constant variance.The random processes that generated the observed series can then be mathematically defined.The residuals left over after this model is fitted should be lI white noisell (completely random)if the model is adequate. Time series can be passed through a mathematical fi 1ter which changes the form of the input series.A Ill ow pass filter ll dampens high frequency perturbations and allows low frequency perturbations to pass unchanged. This is useful in smoothing noisy time series so that the basic pattern may be more readily observed.High pass filters are used when it is desirable to remove obvious (low frequency)trends in order to focus on the high frequency events. Box-Jenkins models can be constructed using only the information con- tained in the time series itself.For example,although the discharge time series results from several independent variables including rain- fall,air temperature,and solar insolation on the glaciers,it is not necessary to quantify these inputs in order to model the output (dis- charge).Information on the effects of all the inputs is already contained in the past history of the discharge record.However,infor- mation on the input series can be used in a transfer function model to obtain an equation with more predictive power.This is a model where an output seri es is a functi on of one or more independent input seri es as well as its own past history. An observed series is one realization of all possible time series which could have been generated from a random process.Time series analysis examines the nature of the probablistic process that generated the 4 - - ~. I ~, - ..... - - - - ,.... - ..- observed series.The model should have similar properties to the generat"ing mechanisms of the stochastic process (Granger and Newbold 1977).Then!one can form summary statistics about the series and make inferences about the nature of the stochastic process.After a model has been developed!it can be used to test some hypothesis about the generating mechani sm of the time series!to forecast future values of the series!or to make decisions on how to control future values of the series (Granger and Newbold 1977). 1.2 Applications of Time Series Analysis Time series analysis has been extensively used in examining physical data!particularly in oceanography.Salas and Smith (1981)demonstrated that ARIMA models can be used to model the time series of annual flows in streams.Srikanthan et al.(1983)analyzed the time series of annual flows in 156 streams in Australia.Time series models have also been used to examine the effect of the Aswan dam on the discharge of the Nile River and the effect ofa hydroelectric dam on the discharge regime of the Saskatchewan River (Hipel et al.1978). Time series analysis methods have been also been used in examining time series of abundance and catch in marine fisheries (Van Winkle et al. 1979;Botsford et al.1982;Peterman and Wong 1984;and Taylor and Prochaska 1984).These methods have been used by Saila et al.1980, Mendelssohn 1981!Stocker and Hilborn (1981),Kirkley et al.(1982),and Jensen (1985)for forecasting future abundance or catch of marine fi sh stocks.Mendelssohn (1981)used transfer function models in addition to univariate Box-Jenkins models to forecast fish catch.Botsford et al. (1982)focused on searching for causal mechanisms of observed cycles in salmon fisheries in California rather than on defining models for the fisheries. Applications to freshwater fish ecology problems are much more limited. Saila et al.(1972)used time series methods to cross correlate upstream migration activity of the alewife to solar radiation and water tempera- ture.O'Heeron and Ellis (1975)considered a time series model for judging the'effects of reservoir management on fish.Applications of spectral analysis to ecological problems have been reviewed by Platt and Denman (1975)and time series analysis in ecology was the subject of a .symposium proceedings edited by Shugart (1978). 1.3 Objectives The objective of this paper was to develop mathemati ca 1 model s for the times series of mean daily Susitna River discharge at the Gold Creek gaging station (river mile 136.7),daily turbidity level!and daily outmigration rates of chinook salmon (Oncorhynchus tshawytscha)and sockeye salmon (0.nerka)at the Talkeetna Station outmigrant traps (river mile 103.0)during the open water seasons of 1983 and 1984. Because time series analysis can provide an efficient summarization of a data set by a few parameters (Hipel et al.1978),these models will be used to statistically describe the present conditions as a baseline against which future changes can be measured.The discharge and tur- bidity information will be useful for examining their relationship with 5 salmon fry outmigration as well as with other species and life history stages.In addition,discharge was used as an input in transfer func- tion models of discharge-turbidity,discharge-chinook outmigration and discharge-sockeye outmigration in order to describe the relationship between these variable and to be used as a possible technique to fore- cast futu re values or to exami ne the probable effects of the proposed dams. Turbidity was chosen as a variable of interest because of its rela- tionship with discharge and because of its importance in determining the distribution of rearing juvenile salmon (Suchanek et al.1984;Part 2 of this report)and other species.It was selected more for this reason than for its effect on salmon outmigration,so it was not used as an input in a transfer function model with salmon outmigration.Chinook salmon were chosen because this species rears in sloughs and side channels affected by mainstem discharge and because chinook salmon have been selected as the evaluation species of the impact assessment study (EWT&A 1985).The sockeye salmon time series was chosen because mainstem discharge affects sloughs which are both natal and rearing areas for this species.While chinook salmon spawn mainly in tributaries in this system,sockeye salmon spawn mostly in mainstem sloughs. 6 - ~, - ~I .- ,-, , - /"''' 2.0 METHODS 2.1.The Data Mean daily discharge values for 1983 and 1984 (Fig.2,Fig.3)were obtained from the U.S.Geological Survey gaging station on the Susitna River at Gold Creek,river mile 136.7 (Still et a1.1984;U•.S.Geolog- ical Survey provisional data,1984).The time frame examined was May 18 to August 30 (105 observations).Discharge levels begin to decline in September when glacier melting decreases;hence,a .10nger series would not be stationary.Throughout this paper,the unit for discharge is one thousand cubic feet per second. Daily water samples for turbidity (Fig.2,Fig.3)were taken at the outmigrant trap station and measured with an HF Instruments Model No. ORT-15S field turbidometer (Roth et al.1984).Units are in nephelo- metric turbidity units (NTU).Only the 1984 turbidity series was examined.. Outmigration rate (Fig.2,Fig.3)was measured by two outmigrant traps, one on each bank,located at river mile 103.0 (Roth et a1.1984).The rate·is reported as number of fish per trap hour with catch from the two traps combined.Only age 0+fry were used in the analysis because the traps were not efficient at capturing age 1+fry and,consequently,the numbers were low.Further,age 1+chinook and sockeye salmon have essentially completed their outmigration from this reach of river by the end of July so the time series are shorter. The chinook salmon time series for.1983 runs from May 18 (shortly after ice-out)to August 30 (when outmigration is winding down),a total of 105 observations.The 1983 sockeye salmon data were not examined. There were six days during the 105 day series when the outmigrant traps were not fished - a one day,a two day,and a three day period.Although values for gaps in time series can be estimated by a spl ine method,the gaps in the outmigration series are short enough so that a s"imple interpolation of values is sufficient (Sturges 1983). In 1984,the traps were continuously operated from May 14 to October 6. However,the series were cut off at the end of August in order to be comparable to 1983 and to achieve a stationary series.About 98%of the cumulativeoutmigration of age 0+chinook and sockeye fry in 1984 had occurred by the end of August. 2.2.Identification and Estimation of Time Series Models Univariate models were developed for the four time series:discharge, turbidity,and chinook and sockeye salmon outmigration.Methods for developing Box-Jenkins ARIMA and transfer function models are described in section 7.0.Basically,there are three steps in developing an ARIMA model:model identification,parameter estimation,and diagnostic checking (Box and Jenkins 1976).The autocorrelation (AC)and partial autocorrelation (PAC)plots for each series were examined to help identify possible autoregressive (AR)and moving average (MA)com- ponents.A tentative model was developed and the parameters estimated. 7 - ~- ..... - 150 SUSITNA RIVER DISCHARGE !lQ !000 J.;;. 8.;301ia:e. ~20~u '0 0 ......,Juft 1!5 .luI ,oIu.1:5.Au<:!,_l!5 1983 ~ SUSITNA RIVER TURBIDITY 400 ~300 ~ .ili!200~ '00 0 Jun ,......'5 .suI ,Jul lS Au<:!1 _'5 1983 ... AGE 0+CHINOOK"•• '5 14 13 '2~":I:'0,gQ. '"II~7u IS 5 4 3 2 • 0 .Iu~•Juft 1!5 .Iu',""",r 1:5 1983 ... .7 AGE 0+SOCKEYE'.'IS .4 '"'2~"'"'019 ~IS ;;, u .... 4 "2, 0 '9113 Figure 2.Discharge,turbidity,and chinook and sockeye salmon outmigration rate,1983. 8 so,.---=::"='==~-=~=:---:-:~...., SU,SITNA RIVER DISCHARGE 50 '0 ,,.,.., - o -In.'''''''rmm......'''''''''''''''''"'''""'''''''mmmm'''''''"""'mmmmmm.......l ,Ain 1 Jun 1~Jul 1 ,Jul 15 AU9'Au9'~ loge. """.-------------~~--'----.....SUSITNA RIVER T..,RBIDITY - 100 o-lm.'""'mmrmm......mmrmm"""'rmm......mm"""'rmm,;,.",mm......~ .Jun 1 Ju"'5 Jut 1 Jul'5 .Auq 1 Au9 15 '984 oJun 1 Jun 1.$.Jut 1 Jul'e Au9'AUtJ'5 198'4 O-irmrrrmrrrmmmnrnTmmmmmmrmmrmm""""rmmrmm""""",""""",rmm,,,..l .I!!.----:-'="=~-..".,....,--:----'-----..., u AGE 0+,CHINOOKI. 1ll... 1.3 12 10 10 1I a 7 "II.. .3 2 - II!!,.------__--------...., 17 .~GE 0+SOCKEYE III 15,.. 1.3 12 tl '0 1I I!! 7•5 4 .3 2, O-M1lmmnitft;;(;","""'rmmmmmmmm"'"""""'rmmmm.....;.~~l'n11fI ...km'Jutt HI ..luI 1 .A.I(1!5 .Auq'~'S 'lIS4 "... Figure 3.Discharge,turbidity,and chinook and sockeye salmon outmigration rate,1984. 9 Insignificant components were removed from the model.The residuals were checked to see if there was significant departure from the assumption that they were white noise.If the residuals were white noise,the model was considered to be adequate.If not,a new model was identified and the process repeated until the residuals were reduced to a white noise process. All of the time series work was done using the BMDP statistical package (Dixon et al.1981).The BMDP Box-Jenkins program estimates parameters by both the conditional least squares method and the backcasting method. The estimates chosen for this paper were from whichever method gave the lowest residual mean square. The time series of mean daily discharge from May 18 to August 30 ap- peared to be stationary so no differencing was done.A plot of the range of sub-groups of the series against the mean of the sub-groups (as suggested by Hoff (1983)indicated that a logarithmic transformation of the data would be helpful in stabilizing the magnitude of the fluctua- tions throughout the series;therefore,a model was also developed for the natural log of the raw data.As the turbidity time series was questionably stationary,models were developed for both the original series and for a differenced series. Models were developed for the chinook and sockeye salmon olJtmigration rate time series on both the raw data and on data transformed by ln (x +1).This transformation was used to avoid taking logarithms of zero;there was zero catch on some days. 2.3 Transfer Function Models Transfer function models (see section 7.0)were developed for discharge/ turbidity,discharge/chinook outmigration,and discharge/sockeye out- migration.Only one input (discharge)was used.Multiple input transfer function models (Liu and Hanssens 1980)or multivariate time series models (Mendelssohn 1982)can be developed,but are substantially more complex. 10 - - - - - - I~ 3.0 RESULTS 3.1.Univariate Model for Mean Daily Discharge The time series of mean daily discharge during the summer of 1983 is shown in Fig.4;the log-transformed data are in Fig.6.Autocorre- lation function (ACF)and partial autocorrelation function (PACF)plots for the raw data are given i nFi g.5 and for the log-transformed data in Fig.7.In all the ACF and PACF plots 5 the "+"symbol on either side of the vertical axis indicates the 95%confidence interval.The first order autoregressive ·component was strong in both the raw and the trans- formed series.The ACF and PACF plots for the raw data indicated that a moving average component was required.Models containing various combi- nations of first and second order AR and MA terms were examined.Of the acceptable models identified 5 the model with the lowest standard errors on the parameter estimates and the least significant residuals was an ARMA(2 52).However 5 the ARMA(I51)was nearly as good as the ARMA (2 52) S05 in keeping with Box and Jenkins'(1976)advice that a parsimonious model (i.e.5 the one with the fewest possible parameters)is desirable, the ARMA(1 51)is considered the "best"model for the non-transformed data.Parameter estimates were: =.992 with std.error of .0135 - h E9,=-.580 with std.error of .0807 The model is: where:Yt is the discharge level at time t and a~is a white noise process at time t Neither the mean nor any of the autocorrelations or partial autocorre- lations of the residuals was significant;therefore 5 the model is considered to be adequate.This equation can be interpretted as:The discharge level for any given day is a function of (the mean 1eve1 5 22.7 cfs,of discharge during the period)plus (most of the previous day's discharge level minus the mean level)minus (about half of the previous day's noise component)plus (the given day's noise component). The plots of both the ACF and PACF on the res i dua 1s from th is model showed a slightly significant spike at a lag of 15 or 16 days.This could indicate that the discharge time series has a periodicity of about 15 days,or sl ightly more than two weeks.Thi s possibi 1i ty was further examined by spectral analysis.The spectrum of discharge (Fig.8)did in 11 Susitna River Discharge.1983 ~ 60 50 "'0 ~C 0 40u ~ ~"'O !.S 30III -~1110~~......u:a 20=s 0 10 ~J a Jun 1 Jun 15 ..lui 1 ..luI 15 Aug 1 Aug 15 1983 Susitna River Discharge.1984 60 50 ~ "'0 C 400 ~-. ..."'0!.~30-~"0.f!§. u:a 20~ 0 .-10 a ~ Jun 1 Jun 15 ..lui 1 ..luI 15 Aug 1 Aug 15 1984 Figure 4.Susitna River discharge time series at the Gold Creek gaging station,1983 and 1984. 12 13 12 11.5 11 '10.5...... 0>..c.u..10is..., 3 9.5 9 a.5 8 Log-transformed discharge.1983 - Jun 1 Jun 15 Jul 1 Jul 15 Aug 1 Aug 15 1983 ~ LOG-TRANSFORMED DISCHARGE.1 984 12..,.....-----------------------, 11.5 11 ......10.5t&J Cl0: !10<.J !!! 0...,z 9.5....I 9 8.5 8 Jun 1 Jun 15 Jul 1 Jul 15 1984 Aug 1 Aug 15 Figure 6.Log-transformed discharge time series,1983 and 1984. - 14 AUTOCORRELAT ION S •1.0 -.8 ••6 •.4 -.2 .0 .2 .4 •S ••1.0 LAG t + t to ".,..++- +·t ++ ++..t ,~++..t +t +......+-...~...... ++ ++ +......... +++... ++ +... ++ +......... ~.. ++ -++ ~... 50 +... +......... r- ! PARTIAL AUTOCORRELATIONS .1.0 -.8 -..-.4 -.2 .0 .2 .4 .S o •1.0 LA. .....,7iI'I'Rr\ -+..... -++.......... .........+ +... 10 ...... ++ ++...+ ++ III ...+++"'-...... t-+++ 20 ++ +++......... +... 25 ++ +... + + 30 + + ,.1lJlII.'!!Il + Figure 7.Plots of autocorrelations and partial ~autocorrelations for 1983 log-trans- formed discharge time series. 15 ..... O"l 2.4 1.8.., :E ;:) a::1.2.... () W 0-0.6 t/) CD 0.00 ..J -0.6 + ... ++ ++++ +++ ++ + ++++++t ++t +++ +++++ ++++-t ++ ++++++++++++,+ I , ,8 , .00 .05 .10 I • ,,, I , .15 .20 .25 .30 FREQUENCY , .35 , I I .40 .45 I .50 Figure 8.Spectrum of 1983 discharge time series. ],J J ~J )J t ~J I J J ))J ~j fact indicate apeak at a frequency of .065 (a period of 15 days).It is not known at this time if this periodicity is urea"'.It may be related to weather patterns in the basin which control solar insolation (cloud cover)and rainfall.A much longer time series of discharge would have to be examined to answer this question.A periodic term could be added to the ARMA(l,l)model (Box and Jenkins 1976)but,given the low signi- ficance level of the periodicity,it does not seem appropriate at this stage of model development. Carrying the idea of parsimony a step further,it can be seen that an ARMA(I,O)model using the log-transformed data is adequate and has the lowest number of parameters.The parameter estimates for this model were: A ¢l =.994 with std.error of <.00005 giving a •qq (~I"tIt _\-10.0)+4.+Nw\1t =10.0 +d \; A A The parameter ¢,was very close to unity.If ¢.were equal to 1.000,the model would be reduced to a random walk model (Chatfield 1984).That is, the log of the discharge for today is the same as the log of the dis- charge for yesterday pl us a random error term.When q;1 approaches 1.000 in a model with only one AR term,the series could be non-stationary (Hoff 1983).To test this,the series was differenced.The residuals from an ARIMA(l,1 ,0)model showed significant spikes,so the differenc- ing did not help;the ARIMA(l,O,O)model is better. The AC's on the residuals of the ARMA(l,O)model were a little better than those of the ARMA(l,l)on the non-transformed data.However,the mean of the residuals was slightly significant,so the ARMA(l,l)model on the raw data is probably superior to this one. The 1984 discharge time series is shown in Fig.4 and Fig.6.The ACF and PACF plots (Fig.9)were similar to those of 1983.An ARMA(l,l) model on the 1984 raw data ~?s adequate,as it was in 1983.Para~ter estimates were:y =23.2;¢>,=.808 (std.error =.0638);and 6,= -.692 (std.error =.0750).An AR(l)model on the log-transformed data was also adequate but,again,had a slightly significant mean residual. The ACF and PACF plots,using log-transformed data (Fig.10),were similar to those of 1983,but perhaPas showed less indication of a moving average process.The estimate for ~,was .994 (exactly the same as the 1983 data),with a standard error of 0.0001,and the estimate for y was 10.0. 3.2.Univariate Model for Turbidity The time series for turbidity in 1983 (Fig.11)was more complex than that of discharge.The ACF and PACF plots (Fig.12)indicated a strong AR(l)component.However,AR{l),AR(2),and ARMA(l,l)models were not adequate to explain the series. 17 .'i'_4 AWi ""'" AUTOCOR RE LATIONS -1.0 -..-.S -."-.2 .0 .2 ...•S ••1.0 LAG t ++t ++-- t +-~,; ++ t + +t +t +t 10 t + +t of'....+..+ 15 +... ++ ++ +... ++ 20 1-+ +t..... +... t +-25 +.....+ ++..i'.... +...~50 ..;...-+ "'"' PARTIAL AUTOCORRELATIONS -1.0 -.'a ••a .4 •.2 .0 .2 •of ••••1.0 ..+ --+.. 1-..+ 1-.. ++ ++ ++ ++ t +..+•+ ++.... ++ t + +•............+..... ++ +t++..1-..1- t + +l'...1-..... +.. ++ Figure 9.Plots of autocorrelations and partial autocorrelations for 1984 discharge time series. 18 AUTOCORRELATIONS -1.0 -.8 -..-.4 -.2 .0 .2 ••••••1.0 LAG ++ ++ ++-- +..-...t ++ ++ ++++....... ++++++~+... .++ ++...... ++ ++...... ++ ++ +... ++ +to ++ ++ ++-++ +......+ ++ PARTIAL AUTOCORRELATIONS -1.0 -.S -..-.4 -.2 .0 .2 .4 ••••1.0 LAG,- +•-++..+ ++ ~.............+..+...+..-+......+..+ ++ to ... t .....to....+..+...+++~++I I +...I ..++++1"~+.....+..... +... 30 +.....+..... .....Figure 10 .Plots of autocorrelations and partial autocorrelations for 1984 log-trans- formed discharge time series . .- 19 Susitna Riv~r Turbidity,1983 soo ,...-------------------------, 400 ~ "~300 ...., ;;. ~ ..Q...200~ 100 Figure 11.Turbidity time series at Tal keetna Station,1983 and 1984. 20 AUTOCORRELATIONS .--1.0 -..-.s -.4 -.2 .0 .2 .4 .S .S 1.0 LA. ..+ ++.-+...... +.-++- ++--+.. t ....... +t.... """..+..+...+....•.. ~..+..... +•+.....+ +.. .~..............+ ++ ++...+...... +....... The series appears to border on being non-stationary because it in- creases in the spring as glacier melt increases and then declines in the fall.(This series would certainly be non-stationary over a longer time frame because the turbidity level is very low in the winter).The slow decay of the autocorrelations in the ACF (Fig.12)also indicated non-stationarity. Further investigation using the raw data showed that the series had a significant second order MA term,while the first order MA term was not significant.Both first and second order AR terms were significant.This gives the model:. ~t::r'1(,.1 t I qCof (l"t-r -It~.r)+.0<'0 (1t-.l -It"./) + ..2 3 2tt-~+2Lt. A with std.errors:on ~,=.0122 J\ on tJ2.=.0234 1\on e =.0988 ~ Note that even though the same notation is used,the white noise process (~~)here is different from that in section 3.1. While this ARMA model is adequate for the time frame examined,in general,an integrated model (i .e.,one with a differencing operation) is probably more appropriate because of the suspected non-stationarity of the raw data.The differenced series (Fig.13),which represents consecutive changes in the original series values,is clearly stationary with a mean close to zero.The ACF and PACF plots for the differenced series (Fig.14)showed that the differenced series could be adequately mode 1ed with just the second order MA term;the fi rst order autore- gression term was not significant in the differenced series.The equation is: ..... ~, •.23 ~t-~t w 'hu"e ~Z t::'#Iv ~-""'t-If ..d-I I\.with std.error on el\=.0972 and the mean of the residuals insignifi- cant.<.- 3.3.Univariate Model for Age 0+Chinook Salmon Outmigration The time frame chosen for Age 0+chinook salmon was the same as that of discharge (Fig.15).The plots of the ACF and the PACF for 1983 (Fig. 16)showed a strong first order autoregresssive component.In fact,an ARMA(l,O)model,abbrevi ated as AR(l),adequately represents the data. Although the plot of the range of sub-groups against the mean of the - 22 1 ]1 1 )l l -I t l J ,t,~"J ORIGINAL DATA +... ++++++...+ ++... ++++++..t++ + + ++ ++...+.+ +++ ++..........+ 1- '+ ... ++++ + ++++ +...'"...'"++++++ ++++.+...+ ++ ++ ... +++++++ +...++++..+..++ ++++++++ ...0 >-....uo o CD 0:110 :::>.... :::> l-z a.o- - o I I I I 18 MAY JUN JUL AUG N W S leo t- Z FIRST DIFFERENCES ... + ......+ + +++......... + + • "t .. ...++ .. +...++++++++++ +++... ++ + ... + + ... + + .. + + + .. + ..... ++ + +. ... ++ + ..... + + ++..+ ...+..+t +t ++..+ + + ++ ... + ...+...++++..++t-+ .0>-t- O -0.0CD 0: ~ t--80 - + I. MAY JUN JUL AUG Figure 13.Differenced turbidity time series,1983. AUTOCORRELAT10NS """',-1.0 -.8 -.IS -.4 -.2 .0 .2 .4 ••••1.0 LAG ~+-+~ ~+++++++..+"""++ ++ ++ +.. +......++-++..+........ ++""'"+.. 20 .... ++ "t +++.... ++ 25 ..+..+.... ++ +T++ ++ PARTIAL AUTOCORRELAT10NS '~ -1.0 -.8 -..-.4 -.2 .0 •2 .4 •• •• 1.0 LAG +..-,-++.... ++ !5 ++++++..+++ 10 ++..+ 4-....+ -+.. +.... +.. ++~,+.... 25 + +++ + 50 .. + Figure 14.Plots of autocorrelations and partial autocorrelations for differenced 1983 turbidity time series.- 24 - Age 0+Chinook Salmon.1983 18 -.--------------------r------, 17 16 15 14- 1.3 12 ·11 10 9 8 7 6 5 4 .3 2 1 o-Hmrr1'm'rfb~rtm:rr:¥rmm'T11TrmmTTTlTl~;:;;m:iTimmm~TTTlTI1lTITI~rr_rl - Jun 1 Jun 15 Jul 1 Jul 15 1983 Aug 1 Aug 1,5 Age 0+Chinook Salmon.1984 Aug 1 Aug 15Jul1Jul15 1984 Jun 1 Jun 15 18.,....-----:---'---. 17 16 15 14 1.3 12 1 1 10 9 B 7 6 5 4 .3 2 1 O-lTrrrrrmmmTTTTTTTl1l1mmT1'lTrnTTTITrmmrnmmTTTTTrrrmrnmT1'lTrmmmmTrmmTTrrrrI ... ~ :I: ba. .cu....ou """ ,- Figure 15.Age 0+chinook salmon outmigration rate time series,1983 and 1984. 25 AUTOCORRELATIONS -1.0 -.'-.8 -.4 -.2 .0 .2 .4 .•.1 1.0 LA.r--I---lf----l--+--+--+--+--t---t--;---r- &0 .. ++ ? + + ++++.... +.. + ++ ++ +..... ++...... +++... + ++---- t---+- ... +... + + ++... +... i-+... + i-.. ++++... +++... ++++ ",.,.. - -I PARTIAL AUTOCORR ELATIONS -1.0 -.'-..-.._.2 .0 .2 .4 .,••1.0LAG.... +i- +... ++.........+ ++...... ++..+ ++++ ++..+++ ++..+..... ++++ ++...... ++++...+ 1-++.........++.. ++++ Figure 16.Plots of autocorrelations and partial autocorrelations for 1983 chinook salmon outmigration time series. 26 - sub-groups indicated the need for a logarithmic transformation,the residual AC's of an AR(1)model on the log-transformed data (Fig.17) were slightly larger (but still insignificant)than those of the AR(l) model on the raw data.The standard error on (DI 'however,was lower with the log-transformed data.ACF and PACF plots for the log- transformed data are shown in Fig.18.The AR(I)model for the raw data is: ·~,("1 t-f -J •5 <)t - ...... ""'"i - "with standard error on ¢,=.0743. TheAR(l)model for the log-transformed data is: "with standard error on ¢,=.0363. The mean of the residuals was not significant. The time series plot for age 0+chinook salmon outmigration in 1984 (Fig.15)shows a different pattern frnm that of 1983.The fry did not begin to migrate in 1984 until about June 12.The low level of out- migration early in the season causes a time series which is non- stationary.To avoid this problem,the time frame selected for 1984 ran from June 12 to August 31 (79 cases).Analysis of this shorter series is not as strong as that of the longer seri es in 1983 but the seri es is long enough from a statistical point of view;Hoff (1983)suggests that about 40 or 50 observations is the minimum necessary for attempting an ARIMA model.Although logarithmic transformation did not appear to be strictly necessary for the 1983 data,it was requi red (to produce an AR(l)model)with the 1984 data,perhaps because of the shorter time series in 1984. The ACF plot for 1984 on the log-transformed data (Fig.19)was similar to that of 1983,although it did decay a little more quickly.The 1984 PACFplot (Fig.19)was very similar to that of 1983 in indicating a strong AR(I)component.The estimated value of¢,in 1984 was 0.973 (very close to that of 1983),with a standard error of 0.0265.The 1984 model is: 27 ,""", Log-transformed chinook,1983 4 3.5 ::5 .......2.5..- + J:2a- 8z ..l 1.5 0.5 04n\-rffrrlirrrmmmTn1Tl1TmmmTnTlTTTmmmTn1Tl1TmmmTnTlTTTmmmTn1Tl1TT1TT111TTT'11 - ..Iun 1 ..Iun 15 ..luI 1 ..luI 15 1983 Aug 1 Aug 15 LOG-TRANSFORMED CHINOOK,1984 -4-,---------. - Aug 1 Aug 15..luI 1 ..lui 15 1984 Jun 1 ..Iun 15 O-fmrnmmmmmrrrmrrfmrrrmmmrnmmmrrrmrrrmmmrnmrnmmmmmmmmmmm-mrrl 0.5 3.5 ::5 --2.5..- + J:2a-u '-'3 1.5 Figure 17.Log-transformed age 0+chinook salmon outmigration rate,1983 and 1984. 28 ,- AUTOCORRELATfONS I"--1.0 -..-.'-.4 -•2 ~.0 .2 .4 ••••1.0 LA. t t +.. to +-- ++- ++ ++ ++++ t + ++ ++ ++++++ ++ ++ ++ ++-++ ++ ++ ++ -++-t -+ 15 +.... ++ ++ +.... ++ ++ ++ ++ 29 15 10 115 20 AUTOCORRELATIONS ·1.0 -..-..-.4 -.1 .0 .1 .4 ••••1.0 + ++ ++-..+ ++...+...t +++....+..+ ++ ++...+ ++ ++ +... +......+..+ ++..++++.. +... ++.......... ++...+ ++ ++ - """ - PARTIAL AUTOCORRELATIONS -i.O -.....-.4 -.2 .0 .2 .4 ••.,1.0 l.AG +... ++ ++..+ 5 ++...+ ++~...+++ 10 ++........."~..... ++ 15 ++...+...+~++ ++ 20 ...+........i...+++ 25 +... +.. ++..+~ ++ 50 ...+ +t+... Figure 19.Plots of autocorrelations and partial autocorrelations for log-transformed 1984 chinook salmon outmigration time series.- - 30 """ .- The mean of the residuals was insignificant.This model does not differ from that of 1983,except that the mean 1eve1 was higher.This was a result of a higher escapement of adult chinook salmon in 1983 than in 1982. All three of the ACF plots for chinook fry outmigration (Figs.16, 18, and 19)had AC's after lag 18 which did not appear to decay further. This may indicate the presence of a weak non-stationary or periodic element which should be explored with subsequent data sets. 3.4.Univar;"ate t10del for Age 0+Sockeye Salmon Outmigration Age 0+sockeye salmon outmigration was examined from May 23 through August 31,1984 (Fig.20).This time series showed a strong AR(l)compo- nent (Fig.21),similar to that of the chinook salmon time series. However,neither an AR(l)model on the raw data or on the log- transformed data was adequate.A MA(l)component was also significant in the raw data,1eadi ng to the model: .... i 1t'::'.1(,t .18 (-d"t .., .- i i - - - ,"'" A ~ The standard error on 'i/J,(.775)was .0681 and on €II (-.567)was .0883. Although the mean of the residuals was slightly signlficant,none of the autocorrelations or partial autocorrelations were,so the model is reasonable . 3.5.Discharge-Turbidity Transfer Function Model The cross correlations for the residuals from the 1983 discharge series and the 1983 turbidity series,both filtered by the ARMA(I,I)model for discharge,had a significant spike at lag =1 day (Fig.22).This suggested a candidate model (Box and Jenkins 1976;McCleary and Hay 1980): ~o 8 1-6.B where:Yt is the output series (turbidity) w 0 and £.are transfer functi on parameters B is the backward shift operator x t is the input series (discharge) Nt is the noise component,an ARIMA model 31 - Age 0+Sockeye Salmon,1983 18 17 16 15 14 13 ~\ 12 ~:::r 110 J:10U9Q. .J:.8u....7Cl 0 6 ~ 5 4 .3 2 ~ 1 .0 Jun 1 Jun 15 Jul 1 Jul 15 Aug 1 Aug 15 -1983 -Age 0+Sockeye Salmon.1984 18 17 16 15 14 13 12 '-:::r 11Q J:10 b 9Q. .c Su..70 -0 ~16 5 4 3 -2 1 0 Jun 1 Jun 15 Jut 1 ..luI 15 Aug 1 Aug 15 ~ 1984 ~ Figure 20.Age 0+sockeye salmon outmigration rate time series,1983 and 1984. .- 32 AUTOCORRELATIONS r-.-1.0 -..-..-.4 -.2 .0 .t .4 ••••1.0 LAB +++.-++ ++ ++ ++ ++ ++ ++ ++ ++ "....++ +-+++ ++ ++ ++ ++ ++ ++ ++ ++ 25 ++ ++ ++++ ++ r""'"SO ++ ++++ ..- i-' i PARTIAL AUTOCORRELAT ION S -1.0 -..-..-.4 -.2 .0 .t .4 ••••1.0 LAG +-to -i +++- ++..+ ++++ ++++++ ++ ++++++++++++++++ ++ ++ +++++++ ++++ ++ + ++++++ Figure 21.Plots of autocorrelations and partial autocorrelations for 1984 sockeye salmon outmigration time series. 33 - CORR ELATIONS ~CROSS .0 .2 .4 .f ,. .1 1.0.1.0 ..I -.s -.4 •.2 LAG ~ •20 +•++ ++ +•+..-.-Ie ++ ++•.. ++ ++ ++ +++'t ++ ++ -15 ..+~,..+++ ++++..+ ++- +++ ++++-+.. ++ +.. ++ ++-.++++ ++..+ ++-..'t ++++ -+.. 20 ....- Figure 22.Plot of cross correlations between the residuals of the ARMA (1,1)discharge model and the prewhitened turbidity time series,1983 data. 34 - - The assumption that the ARIMA component of the model was white noise led to significant AC's in the residuals series and was therefore rejected. The ACF and PACF plots on the residuals from this model suggested an AR(l)model for the Ntcomponent~leading to the full model: OJ o (j J -J"BI + 1-¢,B Parameter estimates were: A 00 0 =8.349 with std.error of 1.7044 -'\ S,=-0.559 with std.error of 0.1718 95,=0.993 with std.error of 0.0009 The t statistic for each of these estimates was significant~leading to the conclusion that discharge and turbidity are related by the equation: t.35 B7t=I+-.5"B /"I t +-.qq B -- The ACF and PACF plots on the residuals from this model showed no significant spikes;therefore~the model is adequate. 3.6.Discharge-Chinook Transfer Function Model After both the input series (discharge)and the output series (chinook salmon outmigration rate)from 1983 were filtered by the ARMA(l~I}model for the discharge series and the residuals from both series were cross co.rrelated~there was a significant correlation at lag =1 day (Fig. 23).This suggested the transfer function model ~as given by McCleary and Hay (1980): or~using the backward shift notation of Box and Jenkins (1976): 35 CROSS CORRELATIONS •1.0 •.8 ~..~.4 ~.2 .0 .2 .4 .S .8 1.0 LAS -20 +t ++ +i- ++++ -15 +.. ++ ++ ++ ++ -10 +-+..+.......+++ + + ++++ ++ ++..+..1'- +.. +...... +........+ +-+ ++.... +of +.. +...... ++ ++..+ +....+....... Figure 23.Plot of cross correlations between the residuals of the ARMA (1,1)discharge model and the prewhitened chinook salmon outmigration time series,1983 data. 36 - - - ~, - This model implies that the current day's discharge rate has an effect on the next day·s outmigration rate.The estimate of W o was 0.02.The residual ACF using this model suggested that the assumption of white noise for the Ntcomponent was not valid;it appeared that an ARMA(I,O) mode 1 wou 1d be appropri ate.The fu 11 model is: + 1-¢,B The parameters for this model were estimated as: .1\ W o =.025 with std.error of .0249 ~l =.667 with std.error of .0751 The t statistic on the estimate for tOo was not significant.However, because the practice of prewhitening the output series with the model for the input series tends to underestimate the si gnifi cance of the results (Botsford et al.1982)and because there was a significant cross correlation between discharge and outmigration rate at a lag of one day, it seemed best to leave this term in the model.This would have to be verified with more years of data.The model is: - - The ACF and PACF for the residuals from this model showed no significant spikes so we may conclude that the model is adequate. This model does not imply that the discharge series is a strong predic- tor for the outmigration series.But adding discharge does result in an expression which has more predictive value than would be obtained by looking at the outmigration series by itself. 3.7.Discharge-Sockeye Transfer Function Model As with the di scharge-chinook rel ationshi p,the cross-correl at;ons of the 1984 discharge and sockeye series,filtered by an ARMA(l,I)model for discharge,showed a significant spike when the sockeye series was lagged one day behind the discharge series (Fig.24).This spike was stronger for sockeye than it was for chinook.A candidate model (Box and Jenkins 1976;McCleary and Hay 1980)was: I ~8 t3,!'It +- 37 CROSS CORRELATIONS •I.0 -.8 -.6 •.4 -.2 .0 .2 .4 .e .8 1.0LAG -20 t +...+...+++ ++ -15 ......+++of- +i- +of- -10 ++ T +...to ++++ -5 ..... +...+... +...-++ 0 ++++- +......+-t + ++++...... ++++ 10 ........+ ++++ +ot-++ +... ++ ++....t- 20 -++ Figure 24.Plot of cross correlations between the residuals of the ARMA (1,1)discharge model and the prewhitened sockeye salmon outmigration time series,1984 data. 38 - - ...., The ACF and PACF plots on the residuals from this model suggested an ARMA(l,l)model for the Nt component,leading to the full model: lUI)B 1-ef,B +((-$1 8) (I-fl.B) - ""'", Parameter estimates were: "-Wo =.206 with std.error <.00005 A 8,=-.190 with std.error .1848 1\ (/>,=.952 with std.error .0483 "e,=-.318 with std.error .1078 A The t statistic for each of these estimates except J was significant, giving:' -3whereIYt=discharge X 10 The ACF and PACF plots on the residual series from this model showed no significant spikes and the mean of the residuals was barely significant; therefore,the model is deemed adequate. -, - ,,-t -(l+.3.2B) ( \ -.15 G)ttt 39 4.0 DISCUSSION I Time series analysis is a useful method for dealing with time ordered data sets,including ones that do not appear to make much sense at first glance because they are too noisy or because they drift as a result of .random events.The·modeling effort helps us to understand why the plots look as they do and what factors shape them.It also is useful in trying to understand what effect a change in the controll ing factors might produce. The influence of discharge level on turbidity and chinook and sockeye salmon outmigration is clearly seen upon inspection of Fig.2 and Fig. 3.Of course,these latter three series are shaped by several factors other than discharge,so the correlation coefficient between them and discharge is not normally expected to be high,unless a relatively short section is examined.For example,the discharge peak in early June of 1983 is reflected in the other three series (Fig.2).The bimodal discharge peak in August of 1983 is reflected in the turbidity and the chinook outmigration series,but only the first August peak is shown by the sockeye outmigration series.Thi s was because most age 0+sockeye salmon in the reach above the traps had left by the middle of August. Similarly,the late August discharge spike in 1984 had no effect on the sockeye series (Fig.3).However,the high discharge peak in mid June of 1984 is strongly reflected in the sockeye series because this was a time when many age 0+sockeye salmon were present in the reach. Another example of a change in the relative effect of a discharge spike is shown by the 1984 chinook salmon series.The high discharge peak is mid-June had less effect on chinook outmigration than did the lower discharge peak in late July,a time when more age 0+chinook fry were ready,because of physiological and behavioral reasons,to outmigrate. The segments of the time series examined (discharge,turbidity,chinook and sockeye salmon outmigration)were described by relatively simple Box-Jenki ns models,usi ng mostly fi rst-order terms.The useful ness of Box-Jenkins models is shown by the relative simplicity of the models developed for the salmon outmigration series;a visual inspection of the plots of the raw data for these series (Figs.15 and 20)gives the impression of an erratic series of events.None of the series appeared to require differencing (although turbidity was on the borderline)to achieve stationarity nor did they appear to have a periodic component (discharge being a possible exception)which would require seasonal differencing.However,this should be re-examined when subsequent seasons of data are available.All of the series showed a strong first order autoregressive term,indicating that the value for anyone day is greatly influenced by the value for the previous day.Similar results for the flow regimes of several streams in Austral ia was reported by Srikanthan et al.(1983),who found that most of the discharge series which were not white noise had a first order autoregressive term. Examination of the autocorrelation coefficients of the four time series at lag =1 day (adjacent values)gives an idea of the smoothness of the time series.Typically,the coefficient for physical/chemical variables is higher than that of biological variables and the time series for 40 - - - - discharge (Fig.4)and turbidity (Fig.11)are less jagged than those for chinook salmon outmigration rate (Fig.15)and sockeye salmon outmigration rate (Fig.20).Saila et al.(1972)reported similar results for the autocorrelations of alewife upstream migration activity in relation to incident solar radiation and water temperature. The square of the autocorrelation coefficient at lag =1 gives a measure of the percentage of the variance of the value for today which is explained bY2 what was measured yesterday (Murray and Farber 1982).In 1983,(.86)=74%of the variance of discharge for one day was explained by the value for diicharge on the previous day.The percen- tage for turbidity was (.92){85%while,for chinook salmon outmi 2grationrate,it was only (.66)=44%,and,for sockeye salmon,(.65) =42%. So,although fish tend to move in pulses more so than water or suspended sediments,fish outmigration is far from being a random event.That is, when an outmigration pul se occurs,the impetus has affected many fish and the phenomenon extends over a three or four day peri ode When we look at an outmigration time series,we are seeing the integrated results of several factors operating on sub-groups of the population in different locales.The fry in one slough may have emerged two weeks earl ier than those of another slough because of a higher intragravel temperature.Or the head of one slough may have overtopped at a lower discharge level than the head of another slough,thus providing an environmental cue to the two groups at different points in time.But there is also a behavioral effect in that fry are stimulated to migrate when they see other fry migrating.This is particularly true for those species that form schools during outmigration. The turbidity time series was the only one examined which included a second order term.The second order moving average term is 1ikely re'lated to the random "s hock"caused by a rising discharge (which is in turn caused by rainfall)which resuspends sediment.It takes a few days after the rainfall is over for this perturbation in turbidity level to drop to the pre-rainfall 1evel. The discharge-turbidity transfer function model does not necessarily imply that discharge level is a strong causal factor for turbidity. These two variables are correlated largely because when glacial melting is high,both discharge and turbidity are high.This phenomenon pro- vides the seasonal trend of the two series;the discharge of clear water tributaries such as Portage Creek and Indian River (which increases discharge but not turbidity)is a noise component.However,discharge does in fact have some direct causal effect on turbidity by resuspending sediments and other particles during a rapid rise in discharge level. Certainly turbidity is not a cause of discharge,so it makes sense to take discharge and noise as the input and turbidity as the output of a transfer function model.The value of the model is that it allows levels of turbidity for a few days ahead to be predicted from past values of both turbidity and discharge. 41 Turbidity level after the dams begin operation will not only be influ- enced by a changed discharge regime,but will also be directly changed by the dams because of settling of suspended sediments in the reservoir. By building Box-Jenkins models for these four time series,a better understanding of the processes·which control these variables was developed in that the structure of the random processes which generate an observed series has now been specified.Also,we have statistically described the natural time series as a baseline against which future changes can be assessed.Thi s descri pt i on of the di scharge and tur- bidity regimes is important not only because of their effects on salmon outmigration,but also because of their effects on other life stages and species.It is important to explore the effect on salmon outmigration of a construction project which will change the basic rules,that is, change the underlying physical processes.Whereas the present discharge regime can be described as a mixed first order autoregressive and moving average process,the di scharge regime under a post-project scenario could include entirely different terms. An important point is that the underlying processes (the autoregressive and moving average components)were essentially the same in 1983 and in 1984 even though the actual time series,or llrealizations,ll looked very different between the two years.This was true for both discharge and for chinook salmon outmigration;only a single year of turbidity and sockeye salmon outmigration was examined.Even though the discharge peaks do not match between the two years and the mean levels between years may have been different,the process which generated these peaks in both years was the same and can be described by an ARMA(l,l)model with similar parameter estimates for both years. In a sense,the proposed dams would operate like a gigantic low pass filter on the discharge regime,dampening out the high frequency pertur- bations and letting the low frequency (annual cycle)events pass,but at a reduced amplitude.In other words,there are two effects of intro- ducing a reservoir into this system:1)the day-to-day changes in discharge would be smoothed and 2)the general discharge level would be higher than normal in winter and lower than normal in summer.However, this is an oversimplification because a new element would be present if the dams are built -namely,power demand.Power demand is not in phase with the natural discharge fluctuations,so dam operation to accommodate power demand wi 11 change the mechani sms whi ch generate the current discharge regime. The important question is,how would the salmon outmigration rates be affected if these di scha rge spi kes were not present,as wi th a dam- regulated discharge regime?Further,what effects would these changes have on the population survival rate?Relatively high levels of dis- charge,and possibly four or five day peaks,in the late spring and early summer may be necessary to faci 1itate normal outmi grati on timi ng of juvenile salmon.On the other hand,very high discharge levels at this time of year,which occur naturally,may be harmful to juvenile chi nook salmon if these floods di spl ace the fry downstream from what would otherwise be their rearing areas. 42 - ""'" - - - - - - - - - .- .- Time series analysis is a statistical tool which has many potential applications to the Susitna River Aquatic Studies Program.It would be useful to build Box-Jenkins models for the 36 year record of discharge at Gold Creek gaging station.Because this information is continuous,it can be digitized as monthly,weekly,daily,or even hourly means. Turbidity,temperature,and dissolved gas time series could also be modeled in:this manner.Develop"ing time series models for.the proposed post-project di scharge regime to see whether the post-project di scharge regime is also an ARMA(1,l)process would be informative in assessing d~m-relatedeffects.Intervention analysis,which is an extension of Box-Jenkins models concerned with a natural or human caused change to a system,woul d be an appropriate method to use (Box and Ti ao 1975;Hi pel et ale 1978;Thompson et ale 1982).One could determine if the inter- vention (construction of the dams)would have a significant effect on the time series processes.This method has been used to model the effects of the Aswan dam on the Nile River and of the Gardiner dam on the South Saskatchewan River in Canada (Hi pel et ale 1978).Before and after mean levels can not be compared using normal analysis of variance because the observations are serially correlated.. Developing forecast models for the annual return of adult salmon or the annual total number of outmigrants would be an excellent use of time series analysis.The adult salmon return of a particular year is strongly related to the return of the previous year (that is,when catch is high one year,it tends to be high for several years)and there is probably a periodic component based on strong year classes.With such a model,one could predict the size of next year1s adult salmon return,a piece of information whi ch woul d be very useful to both fishery and hydroelectric dam managers.However,the time series of adult salmon return to the Susitna River is not long enough (only seven or eight years of data)to develop Box-Jenkins models.A minimum of about 40 or 50 observations is necessary (McCleary and Hay 1980;Huff 1983), although the method has been applied by Jensen (1985)to fish catch data with as few as 32 observations.The annual abundance of adult chinook and coho salmon in the California marine fishery has been successfully examined with time series analysis by Botsford et aT.(1982)and Peterman and Wong (1984)have looked at sockeye salmon cycles in British Columbia and Bristol Bay.For the present,analysis of salmon time series in the Susitna River will have to be restr·icted to daily rates of a single year . 43 5.0 ACKNOWLEDGEMENTS I thank Kent Roth,who has run the outmigrant operation since its beginning in 1982,and Dana Schmidt, former Project Leader of the Resident and Juvenile Anadromous Fish project,for their valuable discussions on some of the ideas in this report.Allen Bingham,Paul Suchanek,and Dave Bernard also made helpful comments on a draft copy of the report. Much of thi s work was done as a projec.t for a course on time seri es analysis taught by J.Horowitz of the Department of Mathematics and Stati sti cs,University of Massachusetts.Hi s assistance with the time series analysis and review of this paper are appreciated. I am grateful to Mary Ferber of the Al aska Resources Library for con- ducting a computerized literature search on ecological and fisheries appl ications of time series analysis.Drew Crawford and Andy Hoffmann helped compile this report,the figures were drafted by Carol Hepler, and Skeers Word Processing Services did the typing. 44 - .~ - - - - - • I 6.0 LITERATURE CITED Botsford,L.W.,R.D.Methot,Jr.,and J.E.Wilen.1982.Cyclic co- variation in the Cal ifornia king salmon,Oncorhynchus tshawytscha, silver salmon,O.kisutch,and dungeness crab,Cancer magister, fisheries.Fishery Bulletin 80:791-801. Box,G.E.P.,and G.M.Jenkins.1976.Time series analysis.Fore- casting and control.Holden-Day,San Francisco • Box,G.E.P.,and GrC.Tiao.1975.Intervention analysis with applica- tions to economic and environmental problems.Journal of the American Statistical Association 70:70-79. Brannon,E.l.,and E.O.Salo.(eds.).1982.Proceedings of the salmon and trout migratory behavior symosium.June 3-5,1981.University of Washington,Seattle,Washington. Bulmer,M.G.1978.The statistical analysis of the ten year cycle. Pages 141 ..153 in H.H.Shugart,Jr.(ed.).Time Series and Eco- logical Processes.SIAM-SIMS Conf.Sere 5.Society for Industrial and Applied Mathematics,Philadelphia,Pennsylvania. Cederholm,C.J.,and W.J.Scarlett.1982.Seasonal immigrations of juvenile salmonids into four small tributaries of the Clearwater River,Washington,1977 -1981.Pages 98 -110 in LL.Brannonand LO.Salo (eds.).Proceedings of the salmon and trout migratory behavior symposium.June 3-5,1981.University of Washington, Seattle,Washington. Chatfield,C.1984.The analysis of time series:an introduction. Chapman and Hall.london. Congleton,J.L.,S.K.Davis,and S.R.Foley.1982.Distribution, abundance and outmigration timing of chum and chinook salmon fry in the Skagit salt marsh.Pages 153 -163 in E.L.Bra.nnon and E.O. Salo (eds.).Proceedings of the salmon and trout migratory behav- ior symposium.June 3-5,1981.University of Washington,Seattle, Washington. Dixon,W.J.,M.B.Brown,L Engelman,J.W.Frane,M.A.Hill,R.I. Jennrich,and J.D.Toporek.(eds.).1981.BMDP statistical software.1981.University of California Press,Berkely, California. EWT&A.1985.Instream Flow Relationships Report.Volume No.1. Prepared for Harza-Ebasco Susitna Joint Venture by E.Woody Trihey and Associates and Woodward-Clyde Consultants,Anchorage,Alaska. Godin,J-G.Y.1982.Migrations of salmonid fishes during early life history phases:daily and annual timing.Pages 22-50 .i!!.E.l. Brannon and E.O.Salo (eds.).Proceedings of the salmon and trout migratory behavior symposium.June 3-5,1981.University of Washington,Seattle,Washington. 45 Granger,C.W.J.,and P.Newbold.1977.Forecasting economic time series.Academic Press,New York. Grau,E.G.1982.Is the lunar cycle a factor timing the onset of salmon migration?Pages 184-189 in E.L.Brannon and E.O.Salo (eds.). Proceedings of the salmon and trout migratory behavior symposium. June 3-5,1981.University of Washington,Seattle,Washington. Hale,S.S.1983.Habitat relationships of juvenile salmon outmigra- tion.Appendix H "in Synopsis of the 1982 aquatic studies and analysis of fish ana-habitat relationships.Susitna Hydro Aquatic Studies.Alaska Department of Fish and Game,Anchorage,Alaska. Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt, S.S.Hale,D.L.Crawford,and P.M.Suchanek.leds.).1984. Resident and juvenile anadromous fish investigations (May -October 1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska Department of Fish and Game,Anchorage,Alaska. Hartman,W.L.,W.R.Heard,and B.Drucker.1967.Migratory behavior of sockeye salmon fry and smolts.Journal of the Fisheries Research Board of Canada 24:2069-2099. Hipel,K.W.,D.P.Lettenmaier,and A.!.McLeod.1978.Assessment of environmental impacts.Part one:intervention analysis.Environ- mental Management 2:529-535. Hoff,C.1983.A practical guide to Box-Jenkins forecasting. Wadsworth,London. Jensen,A.L.1985.Time series analysis and the forecasting of menhoden catch and CPUE.North American Journal of Fisheries Management 5:78-85. Kirkley,J.E.,M.Pennington,and B.E.Brown.1982.A short-tenn forecasting approach for analyzing the effects of harvesting quotas:application to the Georges Bank yellowtail flounder (Limanda ferruginea)fishery.J.Cons.into Explor.Mer. 40:173-175. Liu,L-M.,and D.M.Hanssens.1980.Identifi cation of multi pl e-i nput transfer function models.BMDP statistical software.Technical Report No.68.Los Angeles. McCleary,R.,and R.A.Hay,Jr.1980.Applied time series analysis for the social sciences.Sage Publications,Beverly Hills,California. Mendelssohn,R.1981.Using Box-Jenkins models to forecast fishery dynamics:identification,est"imation,and checking.Fishery Bulletin 78:887-896. 46 - - - Mendelssohn,R.1982.Environmental influences on fish population dynamics:a multivariate time series approach.Paper presented at a meeting of the American Statistical Association.August,1982. Cincinnati,Ohio. Murray,L.C.,and R.J.Farber.1982.Time series analysis of an historical visibility data base.Atmospheric Environment 16:2299-2308. Nel son,R. ing. 1973.Applied time series analysis for managerial forecast- Holden-Day,San Francisco,California. ,fIi,fOA ~ I Q'Heeron,M.K.,Jr.,and D.B.Ellis.1975.A comprehensive time series model for studying the effects of reservoir management on fish populations.Transactions of the American Fisheries Society 104:591-595. Peterman,R.M.,and F.Y.C.Wong.1984.Cross correlations between reconstructed ocean abundances of Bri stol Bay and British Col umbi a sockeye salmon (Oncorhynchus nerka).Canadian Journal of Fisheries and Aquatic Sciences 41:1814-1824. Platt,T.,and K.L.Denman.1975.Spectral analysis in ecology. Annual Review of Ecology and Systematics 6:189-210. Priestley,M.B.1981.Spectral analysis and time series.Vol 1: univariate series,Vol 2:multivariate series,prediction and control.Academic Press,London. Raymond,H.L.1968.Migration rates of yearling chinook salmon in relation to flows and impoundments in the Columbia and Snake Rivers.Transactions of the American Fisheries Society 97:356-359. Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of juvenile salmon from the Susitna River above the Chul itna River confluence.Part 1 in D.C.Schmidt,S.S.Hale,D.L.Crawford and P.M.Suchanek (eds.).-1984.Resident and juvenile anadromous fish investigations (May -Qctober 1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska Department of Fish and Game. Anchorage,Alaska. Saila,S.B.,1.T.Polgar,D.J.Sheehy,and J.M.Flowers.1972.Corre- lations between alewife activity and environmental variables at a fishway.Transactions of the American Fisheries Society 101:583-594. Saila,S.B.,M.Wigbout,and R.J.lermit.1980.Comparison of some t"ime series models for the analysis of fisheries data.J.Cons. into Explor.Mer.39:44-52. 47 Salas,J.D.,and R.A.Smith.1981.Physical basis of stochastic models of annual flows.Water Resources Research.17:428-430. Shugart,H.H.,Jr.(ed.).1978.Time series and ecological processes. Proceedings of SIAM-SIMS Conference.Society for Industrial and Applied Mathematics,Phil?delphia. Solomon,D.J.1982a.Migrat'ion and dispersion of juvenile brown and sea trout.Pages 136-145 in LL.Brannon and LO.Salo (eds.). Proceedings of the salmon and trout migratory behavior symposium. June 3-5,1981.University of Washington,Seattle,Washington. Solomon,D.J.1982b.Smolt migration in Atlantic salmon (Salmo salar L.)and sea trout (Salmo trutta L.).Pages 196-203 in £oL. Brannon and E.O.Salo (eds.).Proceed"ings of the salmon arld trout migratory behavior symposium.June 3-5,1981.University of Washington,Seattle,Washington. Srikanthan,R.,LA.McMahon,and J.L.Irish.1983.Time series analyses of annual flows of Australian streams.Journal of Hydrol- ogy.66:213-226. Stevens,D.E.,and L.W.Miller.1983.Effects of river flow on abun- dance of young chinook salmon~American shad,longfin smelt,and delta smelt in the Sacramento-San Joaquin River system.North American Journal of Fisheries Management 3:425-437. Still,P.J.,R.D.Lamke~J.E.Vaill,B.B.Bigelow~and J.L.VanMaanen. 1984.Water resources data.Alaska.Water year 1983.U.S.G.S. Water-Data Report AK-83-1.U.S.Geological Survey~Anchorage, Alaska. Stocker,M.,and R.Hilborn.1981.Short-term forecasting in marine fish stocks.Canadian Journal of Fisheries and Aquatic Science. 38:1247-1254.- .. Suchanek,P.M.,R.P.Marshall,S.S.Hale~and D.C.Schmidt.1984. Juvenile salmon rearing suitability criteria.Part 3 in D.C. Schmidt,S.S.Hale,D.L.Crawford~and P.M.Suchanek (eds.)-.1984. Resident and juvenile anadromous fish investigations (May -October 1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska DepQrtment of Fish and Game.Anchorage,Alaska. Taylor,T.G.,and F.J.Prochaska.1984.Incorporating unobserved cyclical stock movements in fishery catch equations:an applica- tion to the Florida blue crab fishery.North American Journal of Fisheries Management 4:67-74. Sturges,W.1983.On interpolating gappy records for time analysis.Journal of Geophysical Research.88:9736-9740. series - - 48 - Thompson,K.W.,M.L.Deaton,R.V.Foutz,J.Cairins,Jr.,and A.C. Hendricks.1982.Application of time series intervention analysis to fish ventilatory response data.Canadian Journal of Fisheries and Aquatic Sciences 39:518-521. Van Winkle,W.,B.L.Kirk,and B.W.Rust.1979.Periodicities in Atlantic Coast striped bass (Marone saxatilis)commercial fisheries data.Journal of the Fisheries Research Board of Canada 36:54-62. 49 7.0 BOX-JENKINS ARIMA AND TRANSFER FUNCTION MODELS Box-Jenkins models can be summarized as follows (Box and Jenkins 1976; McCleary and Hay 1980;Chatfield 1984).Suppose there is a time series Y:t ' t =l..N.Then Y1;is a moving average process of order q (or an MA(q)process)if where 9 4 are constants and eo =l.The term at.is a whi te no;se pro- cess.White noise consists of a series of ranoom shocks,each dis- tributed normally and independently about a zero mean with a constant variance.The series Yt is an autoregressive Rrocess of order p (or an AR(p)process)if ¢.'i t-I +-¢:J.l't-~t-." t where 9;are constants.This is similar to a multiple regression model except that Yt.is regressed not on independent variables but on past values of itself.A first order autoregressive process,AR(1),has the ~ form: t tlt Box and Jenkins (1976)define a backward shift operator Bas: For m =1, or,the previous value. -50 ,~ Using B,the AR(I)equation can be written: Time series resulting from a mixture of ARand MA processes are called ARMA(p,q)models and have the form:-t .. . +t + ,..,.. - 6.2L t _1 t-.'.+e t 2t t -'t Using the backward shift operator B,an ARMA (1,1)may be written as: ARMA (p,q)models are appropriate only when the time series is station- ary.Stationary in an ARMA model means that there is no systematic change in the mean or the variance over time and that there are no strictly periodic variations (Chatfield 1984);in other words,the mean, variance,and autocovaria.nce are not dependent on time.Time series which are not stationary can sometimes be handled by IIdifferencing"the series.Taking the difference of adjacent values gives a differencing order,d,of one: ) Such models are said to be "integrated ll and are denoted by ARIMA(p,d,q) where pis the order of the autoregressive component,d is the order of differencing,and q is the order of the moving average component. Time series with seasonal variations,such as would occur in a multiple year series of daily water temperature measurements,can be made sta- tionary by seasonal differencing.For example,the value for April 15 of one year is subtracted from the value for Apri 1 15 of the fo 11 owi ng year,and so on for all days of the year. It has been assumed above that the time series had a mean value of zero. With stationary time seri es whi ch have a non-zero mean,the mean has to 51 be subtracted from every y i term.For exampl e,the form of an AR{I) model would be: 1't::',+4>I (~t -I -,..)+~t The autocorrelation function plays a major role in identifying and building time series models.A regular correlation coefficient measures the correlation between N pairs of observations on two variables.The autocorrelation coefficient is somewhat similar except that it measures the correlation between all observations of the same variable at a given distance apart in time (that is,between Yt and 'ft.-jc.for all values of t!t where k =time lag).Also,the covariance is estimated only over N-k pairs of observations (McCleary and Hay 1980).Autocorrelation coeffi- cients at different lags indicate the extent to which one value of the series is related to previous values and can be used to evaluate the duration and the degree of the "memoryll of the process.The autocorre- lation function (ACF)is the set of autocorrelation (AC)coefficients at different lags associated with a time series;a plot of the ACF is called a correlogram (Chatfield 1984). The ACF is defined as: (OV'fU-i 4'1({.(Yt )YttAl V Hi'1lo~(e.C.Yt.') and is estimated by: - - - ACr.k.- N •N-Jt A partial autocorrelation (PAC)coefficient measures the excess corre- lation at lag k which is not accounted for by an autoregressive model of order k-l.The set of PAC I S at different 1ags associated with a time series is called the partial autocorrelation function (PACF).~ There are three steps in developing an ARIMA model:model identifica- tion,parameter estimation,and diagnostic checking (Box and Jenkins 1976).ARIMA model building is an iterative process.The first thing to do is to look at a plot of the time series.Time series that are not stationary must be made so by trend removal which can be accompl ished by 52 r- i such methods as differencing the series or by polynomial (or other) regression.Examination of the autocorrelation function (ACF)and the partial autocorrelation function (PACF)of a stationary series helps to identify a possible ARIMA model.The next step is to estimate the parameters of the model and again examine the ACF and PACF plots,this time on the residuals from the model.This process is repeated until the residuals show no significant AC's or PAC's at any lag,which indicates that the residuals consist of only a white noise process. When there is an independent variable which is also a time series,a transfer function model can be developed.This model consists of the transfer function component from the independent variable as well as the ARIMA component (or noise component)from the dependent variable (McCleary and Hay 1980)and can be represented as: where:Yt is the output time series Xt is the input time series f(X t -b )is the transfer function component Ntis the noi se or ARIMA component Transfer function models can be bivariate (when there is one independent variable)or multivariate (more than one independent variable). The steps to take in developing a transfer function model (Box and Jenkins 1976;McCleary and Hay 1980;Dixon et al.1981)are:(1)develop an ARIMA model for the input series,obtaining the pre-whitened "input (residuals),(2)filter the output series by the model for the input series,(3)cross-correlate the residuals from the first two steps,(4) identify the form of the transfer function component from the cross correlation function,(5)assuming the errors are white noise,estimate the values for the parameters,(6)identify an ARIMA model for the residuals,(7)if the ARIMA component is not white noise,combine the ARIMA component with the transfer function comRonent to form a new model,(8)estimate the parameter values,and (9)examine the ACF and PACF plots on the residuals from the new model to see if the model is adequate. 53 PART 2 The Relative Abundance,Distribution,and Instream Flow Relationships of Juvenile Salmon in the Lower Susitna River. r I - THE RELATIVE ABUNDANCE,DISTRIBUTION,AND INSTREAM FLOW RELATIONSHIPS OF JUVENILE SALMON IN THE LOWER SUSITNA RIVER Report No.7,Part 2 by Paul M.Suchanek,Karl J.Kuntz,and John P.McDonell Alaska Department of Fish and Game Susitna River Aquatic Studies Program 620 East 10th Avenue,Suite 302 Anchorage,Alaska 99501 ABSTRACT Juvenile salmon abundance and distribution were studied in the lower Susitna River (below the Chulitna River confluence)and juvenile salmon rearing habitat was modelled at 20 sites within the reach.Chinook, chum,and sockeye salmon juveniles made use of side channels;however, high turbidity limited use of side channels located in the Chulitna River plume.Coho salmon juveniles were found primarily in tributary mouths;sockeye~chinook,and chum salmon also were present in these areas.Sloughs,which were limited in occurrence,were not used heavily by any of the salmon species. Both tributary mouths and side channel/slough sites were modelled using one of two habitat models.At tributary mouths,an.increase in weighted usable area with a rise in mainstem discharge resulted from the forma- tion of backwater areas which led to lower velocities and an expansion of the area and amount of cover inundated.At si de channel s ~chi nook weighted usable area increased after overtopping due to a gain in cover suitability (turbidity)~velocity,and area.The weighted usable area response to a ri se in rna i nstern di scha rge for sockeye and chum salmon juveniles at side channels was also usually positive.Habitat indices at side channels for chinook~chum,and sockeye juveniles at mainstem discharges and side channel flows above the overtopping discharge declined as velocities became unsuitably high.Weighted usable area for these species did not always decline at high discharges,however, because of the compensating effect of a larger surface area. i TABLE OF CONTENTS ABSTRACT ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••".i LIST OF FIGURES....................................................iv LIST OF TABLES.....................................................vii; LIST OF APPENDIX FIGURES...........................................ix LIST OF APPENDIX TABLES............................................xv 1.0 INTRODUCTION.. . •. . . •... . . . . •. •. •. •. ••••. •. •••. ...•. •. . . . . . . •. . . 1 2.0 METHODS •••••IiI.................................................3 2.1 Field Sampling Design ,.......................3 2.1.1 Study locations and selection criteria .•..•.•..••~....•. 2.1.2 Field data collection •.........................•........ 2.1.2.1 Resident Juvenile Habitat (RaHAB)model sites •...... 2.1.2.2 Instream Flow Incremental ~1ethodology (IFIM)sites •. 2 ..1.2.3 Opportunistic sites ••••••••••••••••.•••~•••••.•••••. 3 5 5 9 9 - 2.1.3 Schedule of activities and frequency of sampling..•••.•.10 2.2 Data Analysis -....10 2.2.1 Physical data -........................10 2.2.2 Abundance and distribution..............................11 2.2.3 HabitQt modelling of rearing salmon........•.••.•••••.••11 2.2.3.1 Suitability criteria development....................11 2.2.3.2 Instream Flow Incremental Methodology (IFIM)models.11 2.2.3.3 Resident Juvenile Habitat (RJHAB)models............12 2.2.3.4 Mode-l verification..................................14 3.0 RESULTS •••••~•••••••••. •••. ••. . . . •. ••••••. ••. •••••••. •• . •. . ••.15 3.1 Seasonal,Spatial,and Discharge Related Variations i n Ha bita t.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 3.1.1 Macrohabitat type classifications of study sites........15 3.1.2 Chulitna and Talkeetna River plume influences on turbidity of side channels...........................19 3.1.3 Physical responses of sampling sites to mainstem discharge variations...........................19 3.1.3.1 Area 19 3.1.3.2 Cover...............................................21 ii TABLE OF CONTENTS (Continued) 3.2 Distribution and Abundance of Juvenile Salmon...............25 3.2.1 Chinook salmon •.•.•.•.•........•.•.•.........o..........25 3•2•2 Co h0 sal ma n••••.•••••••••••••II Q ••t.'I • ••• •25 3.2 CI 3 Chum sa 1man •••.••••••••••.•a 0 ••Ii'III •••••••II Q ••••0 CI ••••II • • •28 ~ 3.204 Sockeye salmon •••...••Go ••••••••••••0 0 •••••••••••••••GI...28 3.3 Habitat Modelling of Rearing Juvenile Salmon................33 I 3.3.1 Chinook salmon ee e o a.........39 3.3.2 Coho salnlon a 0 ••••••••••••••••e ••••••"...42 3.3.3 Chum salmon ••••.••••••..••.•••.••ecoe ••"'••••••••••••••o...48 3.3.4 Sockeye sa 1man ••••••••.•••co II e 1&III GI a II a 0 ••e ....eo.II •GI •II •••co ••54 4.0 DISCUSSION ••••••••••••••••..•••••••••s~••••o •••oog.e..••••••••64 4.1 Chi nook Sa 1man •••III ••••••.,•••••e e G co ..1:1 e •D •••G ea •••••1:1 Cit .....Q • ••64 4.2 Coho Salmon co ••.,0 a ••Q ••a.GI ••.,•••e 'II e •••••co Cl •e g ••••"•Q a ••II!Io .,• • ••65 4.4 Sockeye Salmon ••••••••••••••e ...e.".10 0 co 8 •••.,••ill ••e 8 ill).,p ••0 CI 0..68 5.0 CONTRIBUTORS •••••••••••••••••••o ••••••e ••••e.ee.e.coo •••080.e.,.70 6.0 ACKNOWLEDGEMENTS ••••••••e e ••••••••••••D •••Q •Cl •Cl •e ••0 0 ••••"•-II • •72 7.0 LITERATURE CITED .•••••••••••••.•••••.•••.,D-.o ••ClCl-.co •••oo ...o ••"o.73 8.0 APPENDICES Appendix A Lower Susitna River Juvenile Salmon Rearing Suitability Criteria Appendix B Modelled Site Turbidities, Juvenile Salmon Catches,Areas, Weighted Usable Areas,and Habitat Indices Appendix C Comparison of the IFIM and RJHAB Modelling Techniques at Two Selected Sites Appendix 0 Hydraulic Models for Use in Assessing the Rearing Habitat of Juvenile Salmon in Six Side Channels of the Lower Susitna River iii - - "'.. LIST OF FIGURES Figure Title Page -1 Location of study sites on the lower Susitna River at which juvenile salmon habitat was modelled,June through October 1984....................4 2 Arrangement of transects and sampl ing cell s within a grid at a hypothetical RJHAB model- ling site 6 ..... 3 4 Turbidities at modelled side channels and tributary mouths on the lower Susitna River, June through October 1984..............................18 Compari son of turbi di ti es in the lower Susitna River below the Chulitna and Tal- keetna River confluences on July 19 and August-16.,.1984 .•............•...•.•...........•....•.•20 5 Area within modelled tributary mouths as a function of mainstern discharge at the USGS Sunshine gaging station,1984..........................22 - - ,..... 6 7 8 9 10 11 Area within modelled sloughs and side chan- nels as a function of mainstem discharge at the USGS Sunshine gaging station,1984.................23 Instream cover response at Beaver Dam Slough, Rolly Creek,and Caswell Creek mouths as a fun~ti on of rna i nstem di scharge at the USGS Sunshine gaging station,1984 •..••.•.•....•...•....•..•24 Seasonal distribution and relative abundance of juvenile chinook salmon on the lower Susitna River,June through mid-October 1984 26 Juvenile chinook salmon mean catch per cell at si de channel s and tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984 .•....•...............•.....••.27 Juvenile chinook salmon mean catch per cell at modelled side channels on the lower Susitna River by turbidity increment,June through mid-October 1984 ...•......•.•...•.......•..•...27 Seasonal distribution and relative abundance of juvenile coho salmon on the lower Susitna River,June through mid-October 1984 29 iv LIST OF FIGURES (Continued) Fi gure Title - -- 12 Juvenile coho salmon mean catch per cell at four tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984 D •••••••••••o ••••••30 ..... 13 Seasonal distribution and relative abundance .of juvenile chum salmon on the lower Susitna River,June through mid-October 1984 ••••••••.••••••••••31 14 15 Juvenile chum salmon mean catch per cell at modelled side channels and tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984 ••.•.••••••••...••32 Juvenile chum salmon mean catch per cell at modelled side channels on the lower Susitna River by turbidity increment,June through mid-July 1984 e •••_32 """" - - 20 18 19 16 Seasona 1 di stri buti on and re1 ative abundance of juvenile sockeye salmon on the lower Susitna River,June through mid-October 1984 •••.•...•••34 17 Juveni 1e sockeye salmon mean catch per cell at side channels and tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984 •••••••••••••••••••••••••••••..35 Juveni 1e sockeye salmon mean catch per cell at modelled side channels on the lower Susitna River by turbidity increment (with and wi thout Beaver Dam Si de Channel),June through mid-October 1984 •••••••••.••••..••••••••.•••.••.35 Juvenile sockeye salmon mean catch per cell at Beaver Dam Slough,Beaver Dam Si de Chan- nel,and Rolly Creek Mouth by samp1 ing period,June through mid-October 1984 •••••••••••.••••••36 Weighted usable area for juvenile chinook salmon at the Rolly Creek Mouth,Kroto Slough Head,and Sucker Side Channel study sHes as a function of mainstem discharge,1984 ••••••••••••....•40 21 Wei ghted usab1 e area and habitat i ndi ces for juvenile chinook salmon at tributary mouth study sites as a function of mainstem dis- charge,1984...........................................41 v - - ..... LIST OF FIGURES (Continued) Figure 22 23 24 Title Page Weighted usable area and habitat indices for juvenile chinook salmon at side channell slough study sites as a functfon of mainstem dischar.9'e,1984 -............43 Turbidity adjusted weighted usable area and habitat indices for juvenile chinook salmon at side channel/slough study sites as a function of mainstem discharge,1984 .•....•....•....•..45 Juvenile chinook salmon mean catch per cell versus seasonal mean habitat indices at side channel and tributary mouth modelling sites on the lower Susitna River,1984.......................46 ,""", 26 I~ 27 25 Weighted usable area for juvenile coho salmon at the Caswell Creek,Rolly Creek,and Beaver Dam Slough tributary study sites as a func- tion of mainstem discharge,1984 ...•.•.....•...•.......47 Weighted usable area and habitat indices for juvenile coho salmon at tributary mouth study sites (excluding Birch Creek)as a function of mainstem discharge,1984........•..•....•.•..•••....49 Juvenile coho salmon mean catch per cell versus seasonal mean habitat indices at tributary mouth modelling sites on the lower Susitna River~1984,......•.•.•.•....•,......•.•..•.•....50 28 Weighted usable area for juvenile chum salmon at the Rustic Wilderness and Last Chance Side Channel study sites as a function of mainstem discharge,1984 ......................•.................51 29 Weighted usable area for juvenile chum salmon at the Trapper Creek and Sunset Si de Channel study sites as a function of mainstem dis- ch'.arge,1984...........................................52 - - 30 31 Weighted usable area and habitat indices for juvenile chum salmon at side channel/slough study sites as a function of mainstem dis- charge,1984 53 Turbidity adjusted weighted usable area and habitat indices for juvenile chum salmon at side channel/slough study sites as a function of mainstem discharge,1984............................56 vi 33 32 Figure 37 LIST OF FIGURES (Continued) Title Page Juvenile chum salmon mean catch per cell versus seasonal mean habitat indices at side channel and slough modelling sites on the lower Susitna River,1984 ••...•.•'.~~•.•...•.......•..•.57 Weighted usable area for juvenile sockeye salmon at the Rolly Creek Mouth and Sucker Side Channel study sites as a function of mainstem discharge,1984 .•......•.•.....•.•..••..•.....58 34 Weighted usable area for juvenile sockeye salmon at the Beaver Dam and Sunset Side Channel study sites as a function of mainstem discharge,1984 10................59 35 Weighted usable area and habitat indices for juvenile sockeye salmon at tributary mouth study sites on the lower Susitna River as a function of mainstem discharge,1984 .••.•••••••.••••••.60 36 Weighted usable area and habitat indices for juvenile sockeye salmon at side channel and slough study sites on the lower Susitna River as a function of mainstem discharge,1984 .•.•••••••..••61 Juvenile sockeye salmon mean catch per cell versus seasonal mean habitat indices at side channel and tributary mouth modelling sites on the lower Susitna River,1984 •.....••••••.•.....••••63 vii ~I - ~, LIST OF TABLES Table Title 1 Percent cover and cover type categories................8 2 Partitioning of wetted channel width into s tream ce 1.1s... . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .13 - 3 4 Classifications and habitat characteristics of study sites on the lower Susitna River at which juvenile salmon habitat was modelled, June through October 1984..............................16 Percentages of lower river habitat modelling sites associated with nine cover-type cat- eg,orles................................................17 7 8 - ..... - .- 5 Evaluation of R.JHAB model quality for extrap- olating WUAs over the range of 12,000 to 75,000 cfs as measured at Sunshine gaging s ta t ion ,.1984 .......•.•.•......'..'.....•......~. . . . . . . . . .3~ 6 Discharge ranges of IFIM models at lower Susitna River sites for which hydraulics are rated acceptable,1984 ••••••••••.•••.•••••••.•.•.•••...38 Preliminary juvenile chinook salmon turbidity criteria derived from lower Susitna River side channel distribution data for turbidi- ties greater than 100 NTU •.••••••••••.•••••.•.•••••.•..44 Weighting factors for turbidity by side channel site for analysis of juvenile chinook salmon habitat use,1984...............................44 9 Weighting factors for turbidity by site for analysis of juvenile chum salmon habitat use, 1984 •.•.••.••••.••••••••••••••••.•.•'....................5'5 viii LIST OF APPENDIX FIGURES Appendix Figure A-I A-2 A-3 A-4 A-5 Title Page Mean catch of juvenile chinook salmon per cell by percent cover category (bars)in clear water of the lower Susitna River,1984 and comparison of fitted suitability indices (lines)calculated in 1984 and for the middle Susitna R1-ver,1983 A-7 Mean catch of juvenile chinook salmon per cell by velocity intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 ••••••••••0...............A-8 Mean catch of juvenile chinook salmon per cell by velocity intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for turbid water in the middle Susitna River, 1983 •••••••••.•••••••••••••••••••••••••••••••••••••••••A-8 Comparison of cover type suitability indices for juvenile chinook salmon in clear water calculated from 1984 lower Susitna River distribution data and 1983 middle Susitna River distribution data ••••••••••••••••••••••••••••••••A-9 Cover type suitability indices for juvenile chinook salmon in clear water calculated from 1984 lower Susitna River distribution data •••••••••••••A-I0 .- - - A-6 Mean catch of juveni 1e chi nook salmon per cell by depth intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 •••••••••••••••••••••••••••••A-II 0 A-7 Mean catch of juveni 1e chinook sa.l man per cell by depth intervals (bars)in clear water of the lower Susitna River,1984 •••••••••••••••••••••••A-12 A-8 Mean catch of juvenile chinook salmon per ce 11 by velocity i nterva 1s (bars)in turbi d water of the lower Susitna Ri ver,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 •••••••••••••••••••••••••A-15 ix ~, .- LIST OF APPENDIX FIGURES (Continued) Appendix Fi gure A-9 Title Me.an catch of juvenile chinook salmon per cell by percent cover category (bars)in turbid water of the lower Susitna River,1984 and fitted suitability index (line)calcu- lated for the middle Susitna River,1983 ••••••••••.••••A-IS A-I0 Cover type suitability indices for juvenile chinook salmon in turbid water developed from 1984 lower Susitna River chinook clear water distribution data -.....•.....•.•....•...........A-I? - - - A-ll A-12 A-13 A-14 Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in turbid water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 •••••••••••••••••••••••••A-I? Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in turbid water of the lower Susitna River,1984 •••••••••••••••••A-18 Mean catch of juvenile coho salmon per cell by velocity intervals (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna Riv-er,19·83 _ _ ' '...•.................A-21 Mean catch of juvenile coho salmon per cell by percent cover category (bars)in the lower Susitna River,1984 and comparison of fitted suitabi 1 ity i ndi ces (l i nes)calculated in 1984 and for the middle Susitna River,1983 ••••••••••••A-21 A-IS Comparison of cover type suitability indices for juvenile coho salmon calculated from 1984 lower Susitna River distribution data and 1983 middle Susitna River distribution data ••••••••••••A-22 A-16 Cover type sui tabil ity indices for juveni 1e coho salmon calculated from 1984 lower Susitna River distribution data ••••••••••••••••••••••••A-23 - A-I?Mean ca tch of j uven i 1e coho sa 1mon per cell by depth intervals (bars)in clear water of the lower Susitna River,1984 and fitted su i tabi 1 i ty index (1 i ne)developed for the middle Susitna River,1983 ••••••••••••••.•••••••••••••••A-24 x -------------------------------------- LIST OF APPENDIX FIGURES (Continued) Appendix Figure Title A-18 Proportion of cells with juvenile sockeye salmon present by velocity intervals (bars) in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 and revised in 1984 for the lower river using professional judgement.•. •. •••••. ••••••••. . ••••••••••••••••••. ••••••A-26 A-19 Proportion of cells with juvenile sockeye salmon by percent cover category (bars)in the lower Susitna River,1984 and comparison of fitted suitability indices (lines)cal- culated in 1984 and for the middle Susitna River,1983 ••••••••••••.•••.••••••••••••••••••.•••••••.A-27 - A-20 Comparison of cover type suitability indices for juvenile sockeye salmon calculated from 1984 lower Susitna River distribution data and 1983 middle Susitna River distribution data ••••.•...••••••••••••••••••••••••••.•••••••••••••..A-27 A-21 Proportion of cells with juvenile sockeye salmon present by depth intervals (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 •••••••••••••.•••••••••••••••A-28 A-22 A-23 Proportion of cells with juvenile chum salmon present by velocity intervals (bars)in the lower Susitna River,1984 and fitted suita- bility index (line)developed for the middle Susitna River,1983 .••••••••••••.••••••••••••..••••••••A-30 Proportion of cells with juvenile chum salmon present by percent cover category (bars)in the lower Susitna River,1984 and fitted suitability index (line)calculated for the middle Susitna River,1983 ••••••••••..••.•.•.••••••••••A-31 ..... A-24 Proportion of cells with juvenile chum salmon present by cover types (ba rs)in the lower Susitna River,1984 A-32 A-25 Proportion of cells with juvenile chum salmon present by depth intervals (bars)in the lower Sus itna River,1984 and fi tted suita- bi 1 ity index (l i ne)developed for the mi ddl e Susitna River,1983 A-33 xi Title - LIST OF APPENDIX FIGURES (Continued) Appendix Figure B-1 Weighted usable area for juvenile chinook sa lmon at the Caswell Creek and Beaver Dam tributary study sites·as a function of mainstem discharge ••........•..•..•....................B-12 B-2 Weighted usable.area for juvenile chinook salmon at the Hooligan and Bearbait Side Channel study sites as a function of mainstem discharge ......•.•.•••..........•......................B-13 .- - B-3 B-4 Weighted usable area for juvenile chinook salmon at the Last Chance and Rustic Wilder- ness Side Channel study sites as a function of rna instem di scharge ...•..............................B-14 Weighted usable area for juvenile chinook salmon at the Island Side Channel and Mainstem West Bank study sites as a function of rnainstem d,i-scha-rge .•.•........•...•.........•'8-15 ..... B-5 Weighted usable area for juvenile chinook salmon at the Goose 2 and Circular Side Channel study sites as a function of mainstem discharge B-16 - .-. - B-6 B-7 B-8 B-9 Weighted usable area for juvenile chinook salmon at the Sauna and Beaver Dam Side Channel study sites as a function of mainstem di scha,rge B-17 Weighted usable area for juvenile chinook salmon at the Sunset and Sunrise Side Channel study sites as function of mainstem dis- charge •.•.......••••........•.•.......•.........•...•..B-18 Weighted usable area for juvenile chinook salmon at the Trapper Creek Side Channel study site as a function of mainstem dis- charge ......................•............•.•...........B-19 Weighted usable area for juvenile chum salmon at the Hooligan Side Channel and Kroto Slough Head study sites ad a function of mainstem di scha rge ,_. . . . . . .... . . . . ...8-23 - B-I0 Weighted usable area for juvenile chum salmon at the Bearbait and Island Side Channel study sites as a function of mainstem discharge 8-24 xii LIST OF APPENDIX FIGURES (Continued)-- Appendix Figure Title B-ll Weighted usable area for juvenile chum salmon at the Mainstem West Bank and Goose 2 Si de Channel study sites as a function of mainstem di scharge ..••••...•.••••••......••...•..~••••.•.....•.•B-25 B-12 Weighted usable area for juvenile chum salmon at the Circular and Sauna Side Channel study sites as a function of mainstem discharge •.•.••.•.•....B-26 B-13 Weighted usable area for juvenile chum salmon at the Sucker and Beaver Dam Si de Channel study sites as a function of mainstem dis- charge ...•...••••.•..••.•.•.•••....•••.•.•.•.•.•.•••..•B-27 - 8-14 8-15 Weighted usable area for juvenile chum salmon at the Sunrise Side Channel study site as a function of mainstem discharg~....•••.•.••.•••••.....••B-28 Weighted usable area for juvenile sockeye salmon at the Caswell Creek and Beaver Dam tributary study sites as a function of mainstem discharge .••••.•••.•••...•••.•.•..•...•••••.••B-30 ""'" 8-16 Weighted usable area for juvenile sockeye salmon at Sunrise Side Channel as a function of mainstem discharge •.•••••...•••••.•.••.•.•••••..•••.B-31 C-l C-2 C-3 C-4 Comparison of site areas calculated with the RJHAB and IFIM model 1i ng techniques for the Trapper Creek and Island side channel study sites .••...••••••.••.•••••••••••.•.•.•.••••••.•....••••C-2 Comparison of weighted usable areas calcu- lated with the RJHAB and IFIM modelling -techniques for juvenile chinook and chum salmon at Trapper Creek Side Channel,1984 •...•.•..••.•C-3 Comparison of habitat indices calculated with the RJHAB and IFIM modell ing techniques for juvenile chinook and chum salmon at Trapper Creek Side Channel,1984 .•.••••••.••••••••••.•.•.•.••••C-4 Comparison of weighted usable areas calculat- ed with the RJHAB and IFIM modell ing tech- niques for juvenile chinook and chum salmon at Island Side Channel,1984 ••••••••••.••••....••...•.•C-5 xiii - - - LIST OF APPENDIX FIGURES (Continued) Appendix Figure Title - ,..... ..... - C-5 Comparison of habitat indices calculated with the RJHABand IFIM modelling techniques for juvenile chinook and chum salmon at Island, Side Channel,1984 ••.•.•••••.•.••.~......•.••.•.....••.C-6 xiv LIST OF APPENDIX TABLES Appendix Table Title A-I Percent cover and cover type categories ...•.•....••....A-2 A-2 Kendall correlation coefficients between habitat variables and chinook catch by cell (N=744)for all gear types,in turbid water .•....•.....A-5 A-3 Kendall correlation coefficients between habitat variables and chinook catch by cell (N=396)for all gear types,in clear water •............A-14 A-4 Calculations of turbidity factors for 1984 lower river data .....•.••.•.......•.....••.••....•.....A-16 A-5 Kendall correlation coefficients between habitat variables and coho catch by cell (N=345)in clear water ...•••......•...•••.••.......•.••A-2D A-6 Kendall correlation coefficients between habitat variables and sockeye catch by cell (N=922)-A-26 A-7 Proportional presence of sockeye salmon associated with the composite weighting factor calculated by multiplying velocity and cover suitabilities together .•..•........•••....•••••..A-29 A-8 A-9 A-I0 Kendall correlation coefficients between habitat variables and chum catch by cell (N=249)for all gear types,turbidity below 200 NTU................................................A-3D Proportional presence of chum salmon fry associated with several composite weighting factors.•.•••. .•.•••.. .. •••••••••..••••.•.. .•.•••••.••.A-33 Summary of revisions of 1983 middle river juvenile salmon criteria for use in the lower Susitna River,1984 ••....••.••.•.•••••••.••..•••..•.•••A-35 """" A-ll Suitability indices for juvenile salmon for velocity,depth,and cover in the lower Susitna River,1984 ••...•...•.........•..•••.•.........A-36 B-1 Turbidities within modelled side channels of the lower Susitna River,June through August 1984 .•...................III ••• ••• ••••••••••••••• • • • • • • • •B-3 xv -LIST OF APPENDIX TABLES (Continued) - Appendix Table B-2 B-3 Title Page Catch and catch per ce 11 ,(CPUE)of juvenil e salmon within lower Susitna River sampling si te s,1984............................................B-4 Lengths of RJHAB model sites in the lower Susitna River,1984 .••••••••~•••.•••••••.••••...•••••.•B-5 B-4 - ,.,.. - - Side channel flows at the 15 modelled side channels as a function of mainstem discharge B-6 B-5 Weighted usable areas and habitat indices for juvenile chinook salmon in lower Susitna Rive r mode 1 site s,1984................................B-9 B-6 Weighted usable areas and habitat indices for juvenile coho salmon in lower Susitna River model sites,1984 ...•..~.......•.•••.•.•.•....•........B-20 B-7 Weighted usable area and habitat indices for juvenile chum salmon in the lower Susitna River model sites,1984 ............•••..•....•.........B-20 B~8 Weighted usable areas and habitat indices for juvenile sockeye salmon in lower Susitna River model sites,1984 •...•.•...••..••••......•.•.....B-29 xvi - - - ,- ~, - 1.0 INTRODUCTION The Susitna Ri ver Aquatic Studies Program juveni 1e anadromous di stri- bution and abundance studies initiated during 1981 and 1982 outlined the general distribution patterns of juvenile salmon and their habitat utilization within the Susitna River (ADF&G 1981a,1981b,1983a,1983b). The 1982 studies also investigated the response of selected areas to mainstem discharge changes and demonstrated species differences in the use of "hydraulic zones"(ADF&G 1983c).These zones were subsections of slough and tributary mouth areas.Some zones were affected by mainstem backwater,other zones were above the backwater,and other zones included mixing areas of the mainstem with slough or tributary flow. The relative use of the hydraulic zones by each species of juvenile salmon was analyzed to provide an incremental index of habitat availa- bility at each site for each species.This analysis provided evidence that the relative use by juvenile salmon of these sites was affected by changes in mainstem discharge.Also,the distribution of juvenile salmon suggested certain microhabitat factors within the zone such as turbidity and the amount of instream cover responded to discharge changes at a higher rate than did zone surface area. Studies conducted during the 1983 open-water season concentrated on the instream flow relationships of juvenile salmon in the middle reach of the Susitna River between the Chulitna River confluence and Devil Canyon (Schmidt et al.1984).Suitability criteria for juvenile salmon were developed and these were used in two types of habitat models to model the site-specific response of juvenile salmon habitat to variations in mainstem discharge.Additional information was gathered on juvenile salmon abundance and distribution in the middle reach. The 1983 studies suggested that juvenile chinook salmon made heavy use of mainstem side channels and used the turbid water in these areas as cover.Juvenile coho,chum,and sockeye salmon tended to occupy areas that were less influenced by mainstem flow. In the Susitna River below the Chulitna River confluence (lower river), the braided nature of the river and lower gradient provides large amounts of potential side channel habitat for juvenile salmon.A study plan was formulated,therefore,to examine juvenile salmon distribution and the usability of different morphological components of the lower Susitna River for juvenile salmon during the 1984 open-water season. The results of these studies,which include the responses of rearing juvenile salmon and their habitat within these morphological components to variations in mainstem discharge,are detailed in this paper.These results will be integrated with responses of side channel and slough complex wetted surface areas to variations in mainstem discharge in order to estimate the response of juvenile salmon habitat in the lower river to flow regulation. Large scale aerial mapping of lower Susitna River side channel and slough complex changes in area with variations in mainstem discharge has been done by Ashton and Klinger-Kingsley (1985).Habitat types identi- fied in the mapping included tributaries,tributary mouths,side 1 sloughs~primary side channels~secondary side channels~clearwater areas~and turbid backwaters.Tributaries,tributary mouths t and side sloughs were defi ned as in the mi ddl e ri ver by Kl i nger and Tri hey (1984).Primary side channels have characteristics similar to the mainstem in the middle river and therefore offer little potential habitat for juvenile salmon and are not discussed in this report. Turbi d backwaters are unbreached channel s whi ch contai n turbid water from being overtopped at higher mainstem discharges and therefore are a transitional habitat type between secondary side channels and side sloughs or clearwater areas.Turbid backwaters are not addressed in this report but their habitat values are probably similar to barely breached side channels.Clearwater areas were also not sampled but are thought to have habitat value similar to that of side sloughs. The major emphasis of this report is the evaluation of juvenile salmon use of secondary side channels and their related habitat values.Some of the larger secondary side channels are considered primary side channels at higher mainstem discharges.Juvenile salmon use of tribu- tary mouths and side sloughs was also evaluated.The macrohabitat evaluation data presented here will be integrated with the aerial mapping data contained in Ashton and Klinger-Kingsley (1985)in later reports to formulate the reach-wide response of juvenile salmon habitat to discharge variations. 2 """ - - - - -- - 2.0 METHODS 2.1 Field Sampling Design Three Juvenile Anadromous Habitat Study (JAHS)field crews,composed of two biologists,examined rearing habitats used by juvenile salmon at selected side channels,tributary mouths,sloughs,and mainstem sites of the Susitna River between the Yentna River confluence (RM 28.5)and Chulitna River confluence (RM 98.5).JAHS sampling was conducted from river boats during the open-water season,with helicopter support enl i sted as needed.The crews operated out of camps located on the Susitna River at the Oeshka River (RM 40.6),Sunshine Station (RM 79.0), and Ta 1keetna (RM 97.5). The JAHS field crews sampled three categories of sampling sites.Most of the sampling occurred at Resident Juvenile Habitat (RJHAB)model sites where the response of the site to changes in mainstem discharge was evaluated along with juvenile salmon use of the site.Crews also sampled Instream Flow Incremental Methodology (IFIM)model sites for fish distribution and abundance at which hydrau1 ic habitat model s were developed.The third category of sites,at which further data on fish distribution and habitat were gathered,were known ~s 1I0pportunistic" sites.Further details on specific sampling techniques and methods used in the JAHS studies are given in earlier reports (AOF&G 1984a,1984b). 2.1.1 Study locations and selection criteria The sampling sites modelled were chosen from side channels,tributary mouths,and side sloughs,which met the following basic criteria: The effects of mainstem discharge (stage and flow)on the sites are measurable. The sites are documented or thought to contain potential habitat for rearing juvenile salmon.Sites with extremely high (>3 feet/sec)velocities were assumed to have little value and were not evaluated. C.The sites are accessible by boat at normal mainstem discharges during the open-water season. The 20 sites modelled with RJHAB and IFIM models were distributed between the Yentna River confluence and Talkeetna (Figure 1).Fourteen of the sites were modelled only with the RJHAB model,four with only IFIM models,and two with both RJHAB and IFIM models.Eight of the sites are located within slough or side channel complexes which were picked by R&M Consultants and E.W.Trihey and Associates as representa- tive of lower Susitna River slough or side channel complexes for extra- polation purposes.For purposes of extrapolation,the side channel complex area data need to be integrated with the habitat modelling data by comparing breaching flows and channel size and type between modelled sites and individual channels within the representative complexes. 3 RIVER MODEL SITE MILE RJHA IFIM Trapper Creek S.C.91.6 X X Birch SloughC!88.4 X Sunrise S.C.a 87.0 X Sunset 5.C.°86.9 X. Beaver Dam SloughO 86.3 X Beaver Dam S.C.a 86.3 X Sucker S.C.a 84.5 X Sauna S.C.79.8 X Circular S.C.75.3 X ·Goose 2 S.C.74.8 X Mainst8m West Bank 74.4 X Island S.C.63.2 X X Caswell Creek Mouth °63.0 X Rustic Wilderness S.C.59.5 X Lost Chance 5.C.44.4 X Bear Bait S.C.42.9 X Rolly Creek Mouth 39.0 X Krato Slough Head 36.3 X Eagles Nest S.C.a 36.2 X Hooligan S.C.a 35.2 X a LOCATED WITHIN REPRESENTATIVE SIDE CHANNEL OR SLOUGH COMPLEXES MAPPED BY ASHTON a KLINGER -KINGSLEY (1985). Coolt Inltlt Figure 1.Location of study sites on the lower Susitna River at which juvenile salmon habitats were model1ed,June through October 1984. 4 - - .... .- - - - - - ,..,.. ,...., ,..,.. - - ,..,.. - - Proportionately more sampling effort was expended within smaller side channels in this study because that is where potential habitat is greatest.Only a portion of the habitat modelling sites were selected to occur wi thi n the representative compl exes because further data on distribution of juvenile salmon at locations throughout the lower river were desired. Four of the sites were normally clear-water sloughs or tributary mouths while the other sites were turbid secondary side channels at normal summer flows.Secondary side channels selected for sampling ranged greatly in size,shape,and overtopping discharge.The majority of the habitat model sites selected were secondary side channels because most of the potential habitat for juvenile fish in areas of the lower Susitna River affected by the mainstem is composed of secondary side channels. Primary s'ide channel and mainstem velocities were so high that they were not considered viable habitat. Opportunistic sampling sites were selected by sampling crews as poten- tial habitat which upon sampling might provide for a better analysis of fish abundance and distribution.Sites sampled were more diverse than the RJHAB and IFIM model sites and included areas within alluvial island complexes. 2.1.2 Field data collection 2.1.2.1 Resident Juvenile Habitat (RJHAB)model sites Two types of data were collected at RJHAB model sites.Habitat data were collected for the purpose of modelling the response of the site to changes in mainstem discharge.Fish distribution data were collected for use in verifying the habitat model data,documenting abundance and distribution,and modifying suitability criteria,if necessary.A discussion of the techniques used in the collection of habitat modelling data will be followed by a discussion of methodology used in the col- lection of fish sampling data. Each o~the RJHAB sites was sampled within a grid consisting of a series of transects with associated sampling cells which intersect the channel of the study site at right angles (Figure 2).Grids were located so that water quality within them was uniform and so that they encompassed a variety of habitat types.Survey stakes and orange flagging were used to mark each transect within a grid.Initial measurements within each grid included distances and angles between transect bench marks. Transects were spaced from 50 to 300 feet apart in order to encompass a variety of habitat types within each grid.Aerial photos of all the RJHAB sites showing placement of all transects within each site are presented in Quane et al.(1985). Up to four 6-by-50 foot rectangular sampl ing cells extending upstream from every transect within each grid were characterized by habitat measurements (Figure 2).If the top width of the wetted channel was greater than 42 feet,two of the four cells paralleled both edges of the channel and the third and fourth cells were located parallel to the shoreline cells so as to split the channel into thirds.If the channel 5 ~ J.-,. {;J ~ iTRANSEC\ TRANSECT 2 TRANSECT I Cell Unit Area Sampled - - - Figure 2.Arrangement of transects and sampling cells within a grid at a hypothetical RJHAB modelling site. 6 ..... ..... .- -- ,- - measured 30 to 41 feet in width ·at the transect,there was a cellon each shoreline of the channel and one cell located approximately mid channel.If the wetted edge was 18 to 29 feet in width ,there was one cellon each side of the channel parallel with the bank.If the channel was 1ess than 18 feet in width,there was only one cell . Transects were numbered consecutively beginning with the transect furthest downstream 'withi n the site.Cells were a 1so numbered consecu- tively from right to left looking upriver.If there were less than four cells within a transect,cells were numbered as if the missing cells were present. One or more staff gages were installed by Aquatic Habitat and Instream Flow Project (AH)personnel at each site to document changes in the stage at each site with changes in mainstem discharge.These gages provided an index to the changes in habitat and hydraulic conditions at the site between sampling occasions.AH staff also developed mainstem stage and site flow relationships and mapped the thalweg at selected sites. Habitat modelling data were collected over a broad range of mainstem discharges.Emphasis was placed on data collection at rnainstem dis- charges of 30,000 to 60,000 cfs as measured at the Sunshine USGS gaging station.When staff gage readings and observations indicated that the stage at the site had changed little from a previous sampling occasion, no habitat data were taken • Habitat data taken at each grid on a modelling occasion included the following.At each transect,the distance between the left and right edge of water and the left bank transect marker was measured.If the water quality within the grid or grids was uniform,one measurement of water pH,temperature,conductivity,and dissolved oxygen was taken.A turbidity sample was collected in a 250 ml plastic bottle and stored in a cool dark location for up to two days prior to analysis.Turbidity was measured in nephelometric turbidity units (NTU)with an HF Instru- ments Model No.DRT-15B field turbidometer.If the water quality within the grid appeared to vary because of mixed water sources,additional water quality and turbidity measurements were taken as necessary to describe these within grid variations. In addition to the above measurements,each sampling cell within the grid was characterized by several habitat measurements.A representa- tive depth and velocity were measured by taking one or more point measurements along the midline of each cell.The entire cell was walked so measurements taken were representative . A vel oci ty measurement was taken at 0.6 of the distance from the top of'the water column at one representative location for the entire cell. Additionally,cover type and amount were estimated in each cell and coded into categories (Table 1).Aquatic vegetation was defined as aquatic plants which are normally completely submerged and do not stand upright.Emergent vegetation consisted of plants such as Equisetem sp. which normally are only partially submerged and stand upright.Over- hanging riparian vegetation consisted of vegetation whose roots are 7 submerged only at flood stage and which typically grow in moist or dry soil.Initially,the total amount of cover of all types was estimated for the enti re cell.Next,the primary and secondary cover type was recorded along wi th a percentage of the total for each.Cover was defined as hiding or escape locations for fish less than or equal to 100 mm in total length. Table 1.Percent cover and cover type categories. .... Group # 1 2 3 4 5 6 %Cover 0-5% 6-25% 26-50% 51-75% 76-96% 96-100% Group # 1 2 3 4 5 6 7 8 9 Cover Type No object cover Emergent vegetation Aquatic vegetation Debris or deadfall Overhanging riparian vegetation Undercut banks Gravel (l"to 3"diameter) Rubble (3"to 5"diameter) Cobble (larger than 511 diameter) .... In September,when the water levels in the Susitna River were low,the cover on all the transects within each site was systematically recorded. One person did the systematic cover coding for all the sites so that between site observer bias was minimized.The cover was recorded by distance from the left bank transect marker along the transect line. Fish distribution data were normally collected from a minimum of seven cells with·in each RJHAB site during each sampling occasion.Cells to be sampled were selected randomly by using a random numbers table (ADF&G 1985).If a cell was missing or could not be sampled due to high· velocities or large depths,an additinna1 cell was randomly chosen for samp1 ing.Consequently,the samp1 ing was not totally random.Each cell selected was then sampled for fish with one pass through the entire cell with a backpack e1ectroshocker or beach seine.The gear type used was considered the most efficient for sampling the cell.Typically,beach seines are more efficient in turbid water while e1ectrofishing gear is most efficient in clear water (Dugan et a1.1984).The area of the ce1~ was recorded so that catches in cells with areas different than 300 ft could be adjusted to this standard cell size.Sampling efficiency of e1ectrofishing and beach seining was assumed to be equal. Additional selected cells were occasionally fished at the site if samp1 ing of the random cell s failed to capture many fi sh because the cells had high water velocities.In this case,the sampling crew fished areas which had more suitable water velocities.Areas fished were not limited to cells on the transects.These data were pooled with the randomly selected cell data for analysis. 8 ~, - After each cell was sampled,juvenile salmon captured were identified to speci es and then released.The total 1ength of each of the fi rst 50 fish of each species in each size class was measured in millimeters. If staff gage readings indicated the stage at the site had not changed from a previous sampling period only limited habitat measurements were taken.These included water chemistry data and a turbidity sample. Fish distribution data were taken during each visit to the site,how- ever.Each cell sampled for fish was al so characterized by a represen- tative velocity,depth,and estimate of cover type and abundance. 2.1.2.2 Instream Flow Incremental Methodology (IFIM)sites In addition to the RJHAB model sites,there were also six sites modelled for juvenile fish using the Hinstream flow incremental methodologyn (IFIM)(Bovee 1982).A Summary of this methodology and specific data collection and modelling techniques are presented in Appendix D of this report.All habitat data used in the IFIM models were collected and analyzed by Aquatic Habitat (AH)personnel.Two of the IFIM sites were also modell edwith RJHAB models using the same transects in order to compa re output from the two mode 11 i ng methods.At these two sites,RJ personnel collected the RJHAB and fish distribution data and AH person- nel collected the IFIM data,so the two models were independent. Fish abundance and distribution data were also collected at the other four IFIM model sites.Sampling effort at these sites was secondary in importance to the sampling of the RJHAB sites.Cells were sampled for fish using the transects placed for the IFIM models.Cells were ran- domly sel ected and then sampl ed wi th the same procedures used at RJHAB sites.Cell numbering was the same as that used in the RJHAB studies. The distance from the transect end markers to the cell edge was mea- sured,however,so that the location of the cellon the transect was spec ifi ed.Other data collected at each ce 11 fi shed included amount and type of cover,water depth,and water velocity.Water chemistry mea- surements and a turbidity sample were also taken at a selected location within the site.. 2.1.2.3 Opportunistic sites In addition to the RJHAB and IFIM sites,other sites were sampled for fish as time permitted to gather juvenile abundance and distribution information at a wider variety of sites and to obtain further data for juvenile suitability criteria.Selected 6-by-50 foot cells were sampled for juvenile salmon at opportunistic sites but no permanent grids or transects were marked.Water chemistry was measured at mid-site.If time permitted,each cell sampled for fish was characterized to amount and type of cover,water depth,and water velocity as were cells sampled at RJHAB and IFIM sites. Early in the sampling season,large differences in turbidity.were noted between sites located on the east and west banks of the Susi tna River mainstem below the Chulitna River confluence.In order to better understand the reason for these differences,turbi diti es were taken 9 within the Talkeetna and Chulitna rivers just above their respective confluences with the Susitna and also in the middle Susitna River above its confluence with the Chulitna River.The turbidity measurements were then repeated in the lower Susitna River below the Chulitna River on the left (west)bank channel,center channel,and right (east)bank channel at several locations from RM 92.7 downstream to RM 60.6.Blueline maps detailing the precise sampling locations are available at the Susitna Aquatic Studies office.Two sets of measurements were taken,on July 19 and on August 16.The measurements were recorded within a four hour period on each date.Turbidity samples were taken at least 30 feet off shore near the middle of the channel. 2.1.3 Schedule of activities and frequency of sampling Field sampling trips,lasting approximately 7-10 days,were conducted bimonthly from June through mid-October.Each RJHAB site was sampled for fish on each sampling occasion if fish habitat was present.Habitat data were collected on at least three occasions when staff gage readings or observations suggested a change in the habitat within a site.The collection of habitat data was therefore dependent upon mainstem dis- charge. The IFIM sites were sampled at least once a month during the open-water season.Opportunistic sites were sampled as time permitted and some were only sampled once.Opportunistic sites were sampled mainly in September and early October when many of the RJHAB and IFIM sites were dewatered. 2.2 Data Analysis All fi el d data were recorded on the appropri ate data forms and trans- mitted to the office where the fish distribution data and much of the habitat data were entered into a mainframe computer data base.Data sorts,summary retrievals,and selected computer files were extracted from this data base as needed.Other habitat data were entered directly into basic programs or commercial software on a personal computer. 2.2.1 Physical data Overtopping flows at the study sites were observed or estimated from staff gage measurements and flow observations.Data were grouped into nine half-month sampling periods from early June (June 1 -June 15)to early October (October 1 -October 15).Due to logistical constraints, the actual sampling periods did not always run from the 1st to the 15th and 16th through the end of the month. An index to the amount and type of cover within the RJHAB and IFIM model sites was calculated by totalling the linear feet of all the cover types along the transects at a mainstem discharge within the range of 49,000 to 57,000 cfs.In addition,at Rolly Creek mouth,Caswell Creek mouth, and Beaver Dam Slough,the response of phys i ca 1 cover to changes in mainstem discharge was plotted by totalling the cover along the tran- sects at all measured discharges. 10 """ ...." - ~ , ,.... "'"" ,..... - - The response of RJHAB site wetted areas to ma i nstem di scharge was plotted using a BASIC language geometry program to calculate wetted area at each transect within a site on each modelling occasion.After fitting these points by hand using professional judgement,site areas at 3000 cfs increments were measured on the graphs with a digitizer.The IFG HABTAT program calculated wetted .areas at the six IFIM sites as a function of side channel flow,and these were also plotted us"ing a mainstem discharge-side channel flow ~e1ationship. 2.2.2 Abundance and distribution The same classification of macrohabitats was used to examine differences in fish distribution among the sites as that discussed in Dugan et a1. (1984).The sites were classified as tributary mouths,side sloughs, and side channels.Tributary mouths are sites which are influenced by tributary flows and backwater effects from the mainstem.Side channels are channels whose upstream berms (heads)are breached by the mainstem while side sloughs are channels whose heads are not breached and whose water sources are upwelling,local runoff,or small tributaries.Side sloughs transform to side channels when their heads are breached by the mainstem.Birch Creek Slough was classified as a tributary mouth in 1984 because road building activities in the upper part of the slough closed the head off from the mainstem.Beaver Dam Slough was also classified asa tributary mouth because it only overtops at discharges greater than 80,000 cfs and normally runs clear.Beaver Dam Slough is much more similar to Rolly Creek mouth than to any of the other side sloughs in the lower reach. Catches within cells with areas other than the standard 300 ft 2 were adjusted to correspond to this standard cell area.The analysis was then based on the adjusted mean catch per cell. 2.2.3 Habitat modelling of rearing salmon 2.2.3.1 Suitability criteria development Suitability criteria have been developed to model the response of juvenile salmon habitat to variations in mainstem discharge at sites located in the middle reach of the Susitna River (Suchanek et a1.1984). As habitat data collection techniques used in the lower river in 1984 were similar to those used during 1983,the middle river suitability criteria were compared to the lower river distribution data and mod- ified,if necessary,in Appendix A.The suitability criteria developed in Appendix A are used in all subsequent habitat modelling for the lower river. 2.2.3.2 Instream Flow Incremental Methodology (IFIM)models The IFIM PHABSIM system of computer programs was developed by the U.S. Fish and Wildlife Service as a means of describing the mosaic of phys- ical features of a stream which includes hydraulic variables such as depth and velocity and other features such as substrate or cover (Bovee 1982).A hydraulic model is first calibrated which describes the response of hydraulic variables such as depth and velocity to stream 11 ----------------------- flow (Milhous et al.1981).The HABTAT program is then used to incorpo- rate output from the hydraulic model and substrate data with the suita- bil ity cri teri a to produce estimates of the habi tat potenti a1 (weighted usable area)for a given life stage of a species.Weighted usable area (WUA)is calculated as follows (Bovee 1982): - WUA =Ci,s X Ai where:C., ,s A., =the composite weighting factor (sometimes called the joint preference factor)for cover,velocity, and depth of the cell (i)for the species and life stage (s) =the surface area of the cell Each cell is a small section of the study channel which is bounded by other cells or the shoreline and extends midway between transects.The WUA for the study site at a given discharge was calculated by,totalling a11 the i ndi vi dua 1 cell WUA IS.The compos ite wei ghti ng factor was calculated by multiplying the suitability indices for cover,velocity, and depth of the cell together.WUA'S at each study site were calculat- ed at flows whi ch corresponded to 3,000 cfs increments of rna instem discharge as measured at Sunshine gaging station. Much more detailed descriptions of the IFIM data analysis methods and hydraulic simulation results are presented in Appendix D.Only selected WUA results as a function of mainstem discharge are presented here.All species and site combinations were run and are available on request but space limitations prevent presentation here.Site/species combinations presented were selected on the basis of fish catches at the site. 2.2.3.3 Resident Juvenile Habitat (RJHAB)models The original RJHAB model was designed to calculate weighted usable areas for the habitat within a site Without using hydraulic models (Marshall et al.1984).The model divided the site into shorelfne and mid-channel sections,and calculated weighting factors for cover and velocity for each section which were then multiplied together with area to produce a weighted usable area estimate at each of the discharges measured. The original RJHAe model was greatly modified for the 1984 analyses. These changes were made so that the RJHAB model cal cul ates wei ghted usable areas similarly to the HABTAT program described by Milhous et al. (1981)that is used in IFIM analysis.Also the cover coding has been standardized so that observer variations in rating cover at different discharges do not lead to variations in cover estimates unrelated to changes in wetted area. The current RJHAB model is a spreadsheet developed on commercial soft- ware.Though no hydraulic model is developed,the current RJHAB model 12 """I ~, closely resembles the HABTAT model in its procedures for calculating weighted usable areas within a site.Instead of calculating weighting factors for cover and velocity in shoreline and mid-channel sections on a given sampling occasion as did the original RJHAB model,each site is partitioned into "s tream cellsll each with a unique area,cover type, cover percentage,velocity,and depth.The site weighted usable area (WUA)is ttien the sum of the "s tream cell"WUA's which are calculated by multiplying the area,cover,velocity,and depth suitabilities together. The velocity and depth measurements of the 6'x 50'sampling cells are assumed to represent a much larger stream cell.The wetted surface area between transects was partiti oned into one to four stream cell s depen- dent upon wetted transect width (Table 2). Table 2.Partitioning of wetted channel width into stream cells. - Wetted Channel Width >42 ft 30-41 ft 18-29 ft <18 ft No.of Stream Cell s 4 3 2 1 How Area Partitioned Cellon each shoreline 6 ft in width,two center cells split the difference. Cellon each shoreline 6 ft in width,middle cell is the rest. Each cell with half the width. Entire width. -- Occasionally,islands prevented a simple partitioning of the site but in each case,areas were partitioned so that sampling cells best repre- sented a given stream cell.Once the wetted width of stream cells was partitioned,a computer program written in BASIC was used to calculate the surface area of each stream cellon each sampling occasion.The areas of islands were estimated from width measurements,observations, and sketch maps and then subtracted from the area of each stream cell. Cover suitabilities for each stream cell were calculated with a BASIC program which integrated the standard cover data taken on each transect with the partitioned wetted width of each stream cell.The cover su i tabi 1i ty of each cover type on the stream cell wetted wi dth was averaged with the other cover suitabilities present (proportional to their occurrence)to give an average cover suitability.For example,if the stream cell was 15 feet in width and ten feet of the width was a cover type with a suitability of 0.5 and the other five feet was a cover type with a suitability of 1.0,the average cover suitability for the cell would be :[(10 x 0.5)+(5 x 1.0)]/15 =0.67. The RJHAB spreadsheet then took the stream cell areas and cover suit- abilities,and multiplied these with the depth and velocity suitabil- ities which .it assigned to the sampling cell depth and velocity measure- ments.The products of these calculations (stream cell WUA's)are then totalled to calculate site WUA's for each sampling occasion.Weighted 13 usable areas for chinook salmon in turbid and clear water and chum, coho,and sockeye salmon were all calculated concurrently. Weighted usable areas were plotted over the range of mainstem discharges sampled.Since initial overtopping flows were estimated for each side channel,WUA response was extrapolated in the range around breaching using this information.Habitat indices were calculated by dividing the WUA of the 'site at a given discharge by the site area at the same discharge and these were also plotted.Only selected site and species combinations are presented here,all other WUA calculations are avail- able upon request.Individual sampling cell measurements are also available upon request. In order to compare output from the RJHAB model with that of the IFIM methodology,two sites (Island and Trapper Creek side channels)were modelled with both techniques.Output from both techniques were graphed as a function of mainstem discharge and then correlated with each other at the measured RJHAB discharges. 2.2.3.4 Model verification Fish abundance data were collected at all of the IFIM and RJHAB sites. High mean catches per cell (CPUE's)should reflect high densities of fish within the site.Since WUA on a per site basis reflects the size of a site,WUAisite is not an index to habitat quality of a site.The habitat index calculated by dividing WUA by site area (at any given discharge),however,does reflect site habitat quality,independently of site area. Variations in mainstem discharge cause fluctuations in the habitat value of a given site.Fish populations within a site may not respond immedi- ately to such variations in habitat value but should adjust after a period of time.Over a season,average densities of fish (as expressed by CPUE)should be positively correlated to the average seasonal habitat index if there is a relationship between the two.A test of the signi- ficance of the correlation between mean seasonal habitat indices and mean catch per cell by species was used to verify the habitat modelling efforts. Mean seasonal habitat indices for each site were calculated for each species with the following procedure.Mean daily discharges for each day between May 15 and October 15 were rounded to the nearest 3,000 cfs increment in the range from 12,000 to 75,000 cfs.The season for chum salmon ran from May 15 to July 15.If the discharge was greater than 75,000 cfs,the discharge was assumed to be 75,000 cfs because WUA's were calculated only up to 75,000 cfs.Corresponding WUA's and site areas corresponding to these discharges were then totalled to find the total WUA and site area for the season.The mean seasonal habitat index was then calculated by dividing the total WUA by the total site area. For chinook and chum salmon,WUA's were adjusted by a turbidity factor before the habitat index was calculated.The turbidity factor was calculated by fitting a suitability index from a to 1.0 on the dis- tribution of mean chum and chinook juvenile salmon catch by 50 NTU turbidity increments.Site mean CPUE's were regressed against site habitat indices at each site. 14 ~i - - ..- 3.0 RESULTS 3.1 Seasonal,Spatial,and Discharge Related Variations in Habitat 3.1.1 Macrohabitat type cl assifi cations of study sites All the study sites were classified into one of three macrohabitat types:tributary mouths,side channels,or side sloughs.Classifica- tion and habitat characteristics of the twenty modelled study sites are given in Table 3.Initial breaching discharges for the side channels ranged from approximately 14,000 to 46,000 cfs with flows controlled by the mainstem at least 50%of the time.Channels with input into the tributary mouth sites were never breached at flows less than 54,100 cfs and site flows were controlled by the mainstem less than 5%of the time. Backwater effects were the only effects attributable to mainstem dis- charge at the tributary mouths on all sampling occasions except at Beaver Dam Slo~gh where discharges greater than 75,000 cfs caused the head to overtop and flow to increase through the site.Even at dis- charges greater than 75,000 cfs however,the major effect of rna i nstem discharge on Beaver Dam Slough was a backwater response. The side slough macrohabitat type was not represented by any of the sites when mainstem discharges were highest during the period from late, June through early August.Side slough habitat increased with decreases in mainstem discharges. Major object cover differences among the model 1 ing sites were differ- entiated by macrohabitat type.An index of cover for each site at a discharge of approximately 52,000 cfs (range 45,500 to 58,800 cfs)was calculated for between-site comparisons of cover (Table 4).The per- centage of the site with the primary cover type,submerged aquatic vegetation,varied from 8.5%to 68.5%for the tributary mouths,while none of the si de channel/sloughs had any submerged aquatic vegetation. Substrate in the form of large gravel (1-3 11 diameter)and rubble (3-5 11 diameter)was the primary cover type and averaged 62%of the side channel area whil e these two cover types only covered an average of 14% of the area of tributary mouth sites.The density of cover at tributary mouths was almost three times that of side channels also.Side sloughs, which by definition are unbreached side channels,typically have less object Cover than side channels. Cover,in the form of turbidity was much more frequent within side channels than at tributary mouths and side sloughs.Turbidities were consistently higher in the side channels than in the tributary mouths during the open-water season (Figure 3).A few turbidities of 100 to 150 NTU were recorded at Rolly Creek mouth and Beaver Dam Slough due to rapid increases in mainstem stage which caused turbid water to intrude into the sites,or in the case of Beaver Dam Slough,by a slight over- topping of the channel head by mainstem water.Turbidities within the side sloughs ranged from 1 to 19 NTU with a mean of 5.2 NTU. 15 Table 3.Classifications and habitat characteristics of study sites on the lower Susitna River at which juvenile salmon habitat was modelled,June through October 1984. ...... 01 Site Side Channels (head open)/ Sloughs (head closed) Hooligan Side Channel Eagles Nest Side Channel Kroto Slough Head Bear Bait Side Channel Last Chance Side Channel Rustic Wilderness Side Channel Island Side Channel Mainstem West Bank Goose 2 Side Channel Circular Side Channel Sauna Side Channel Sucker Side Channel Beaver Dam Side Channel Sunset Side Channel Sunrise Side Channel Trapper Creek Side Channel Tributary Mouths Rolly Creek Mouth Caswell Creek Mouth Beaver Dam Slough 2 Birch Creek Slough River Mile 35.2 36.2 36.3 42.9 44.4 59.5 63.2 74.4 74.8 75.3 79.8 84.5 86.3 86.9 87.0 91.6 39.0 63.0 86.3 88.4 Initial Breaching Discharge (cfs) 23,100 14,000 36,000 35,000 (Est.) 22,700 19,000 34,000 19,000 30,000 36,000 37,000 27,500 46,000 31,000 34,300 43,000 75,000+ 54,100 Percent of Time Flow Controlled by 1 Mainstem in 1984 80 94 62 64 (Est.) 79 86 64 86 68 64 62 71 50 68 64 57 o o <5 <5 Non-mainstem Water Sources Pools only Unknown Minor upwelling Pools only Pools only Pools only Major upwell fng Major upwell i ng Minor upwell ing Major upwelling Minor upwelling Minor upwelling Unnamed tributary Major upwelling None Cache Creek Rolly Creek Caswell Creek Unnamed tributary Birch Creek These percentages based on controlling breaching discharges presented in Quane et al.(1985)for the period from May 15 to October 15,1984. 2 A culvert at the head of this slough is frequently blocked and therefore little mainstem water flows into the slough,even if the slough head is breached.The effect of mainstem discharge on this site is minimal for this reason • .,J J t .~I ])!J J J J J J J J , 1 i }1 I 1 1 J J j Table 4.Percentages of lower river habitat modelling sites associated with nine cover-type categories.Percentages are based on the width of transect with each cover type. Cover Index calculated by dividing total cover by total area of site. Percentage of Site With Primary Cover Type River OVerhang.Ccver 1 Discharge No Emergent Aquatic large RI parlan U.C.Density Side Channel s/Sloughs Mile Date (ch)Cover Veg.Veg.Crave I Rubble Cobble Debris Veg.Banks Total (t) Hooligan Side Channel 35.2 7/14 52400 18.9 0.0 0.0 72.0 0.0 0.0 8.5 0.6 0.0 100.0 13.7 Kroto Slough Head 36.3 7/17 49600 56.4 0.0 0.0 8.6 0.0 0.0 33.5 1.6 0.0 100.1 1.8 Bear Bait Side Channel 42.9 7/13 52400 0.0 0.0 0.0 66.8 0.0 0.0 28.1 3.7 1.4 100.0 11.5 last Chance Si de Channe 1 44.4 7/12 54100 23.5 0.0 0.0 63.5 0.0 0.0 12.3 0.8 0.0 100.1 5.9 Rustic Wilderness Side Channel 59.5 8/12 52900 0.0 0.0 0.0 60.9 30.0 0.0 7.8 0.8 0.5 100.0 13.7 Island Side Channel 63.2 7/19 51600 13.4 0.0 0.0 62.0 21.6 0.0 0.0 1.4 1.6 100.0 10.5 Mainstem West Bank 74.4 Extrapolated 54100 1.0 0.4 0.0 43.4 49.3 0.0 2.2 3.4 0.4 100.1 22.7 Coose 2 Side Channel 74.8 7/20 52600 2.0 0.9 0.0 24.3 51.8 13.7 3.5 3.5 0.2 99.9 22.5 Circular Side Channel 75.3 7/24 56600 20.4 0.0 0.0 48.4 21.3 0.0 5.3 4.6 0.1 100.1 9.3 Sauna Side Channel 79.8 7/23 56600 93.4 0.0 0.0 0.0 0.0 0.0 4.3 2.4 0.0 100.1 0.5 Sucker Side Channel 84.5 7/09 55400 80.2 8.4 0.0 6.6 0.0 0.0 3.9 1.0 0.0 100.1 1.1 Beaver Dam Side Channel 86.3 7/08 57100 55.9 0.9 0.0 18.6 5.9 0.0 18.6 0.0 0.0 99.9 1.9 Sunset Side Channel 86.9 7/22 57800 15.0 0.0 0.0 66.8 9.7 0.0 7.7 0.5 0.3 100.0 4.8 Sunrise Side Channel 87.0 7/07 58800 4.0 0.0 0.0 51.4 44.6 0.0 0.0 0.0 0.0 100.0 10.0 Trapper Creek 51 de Channel 91.6 8/19 57200 2.2 0.0 0.0 39.1 58.8 0.0 0.0'0.0 0.0 100.1 12.3 MEAN 25.8 0.7 0.0 42.2 19.5 0.9 9.0 1.6 0.3 100.0 9;5 Tributary Mouths Rolly Creek Mouth 39.0 7111 55100 6.9 25.2 46.2 0.0 0.0 0.0 21.5 0.1 0.0 99.9 24;2 Caswe 11 Creek Mouth 63.0 8/18 45400 2.9 5.3 4B.2 17.6 0.0 0.0 18.4 1.6 6.1 100.1 19;0 Beaver Dam Slough 86.3 7/08 57100 6.8 9.9 68.5 0.0 0.0 0.0 11.1 3.1 0.6 100.0 57;8 Birch Creek Slough 88.4 7/20 52600 36.8 0.5 8.5 29.2 9.0 0.0 13.6 2.2 0.3 100.1 6.3 MEAN 13.4 10.2 42.9 11.7 2.3 0.0 16.2 1.8 1.8 100.0 26.8 ......1 Ccver density Is the average density of object cOver within the site on a percentage basis •'i SIDE CHANNEL TURBIDITIES (MODELLED SITES ONLY) 1000----------=----------------------, 900 800 •MEANIRANGE ::;)700 I- Z 600 500 >- l- e CD Q:400::> I- 300 200 --•O..L----:----r---~-__r--_r_-~--__.__-___,r___-+----J 100 E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLING PERIOD •MEAN IRANGE TRIBUTARY MOUTH TURBIDITIES (MODELLED SITES ONLY) 150 --------------------------_ 140 - 130 120 - 110 - ~100- z 90- >-80- l- e 70- ~60- ::> I-50- 40 - 30 - 20 - 10 -l ~~I t o ....I....------rT---,Tr------.----r---...--1 --i-T ---i-----T1---"r___---l E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLING PERIOD Figure 3.Turbidities of modelled side channels and tributary mouths on the lower Susitna River,June through October 1984. 18 .- 3.1.2 Chulitna and Talkeetna River plume influences on turbidity of side channels Turbidity measurements of the lower Susitna River taken in west bank, mid-channel,and east bank portions of the mainstem indicate that plume influences of the Chulitna and Talkeetna Rivers extend at least 20 to 30 miles downriver (Figure 4).On September 2,turbidities at RM 83.8 ranged from 60 NTU on the east bank,to 77 NTU in 'mid-channel,and 88 NTU on the west bank.West bank turbidities are much higher than on the east bank,because the Chulitna River is three or more times as turbid as the Talkeetna River and middle reach of the Susitna River. A comparison of turbidities at the modelled side channels located above RM 70 also suggests that the plumes have major effects on turbidities downstream.Mean turbidity at lateral side channels located on the west bank (Mainstem West Bank,Sauna S.C.,and Trapper Creek S.C.)during June through late August was 377 NTU.During the same time period, lateral side channels located on the east bank (Goose 2 S.C.,Sunset S.C.,and Beaver Dam,S.C.)had a much lower mean turbidity of 158 NTU. Mean turbidities for all the side channels modelled with the exception of Eagle's Nest Side Channel have been calculated in Appendix Table 8-1. Many more turbi diti es woul d have to be taken to better del i neate the Chulitna River and Talkeetna River plumes.The large east bank clear water tributaries such as Montana Creek and Goose Creek make the differ- ences in turbidity between the east and west banks of the lower river even larger,and confound analysis of the extent of plumes from the Chulitna and Talkeetna rivers. 3.1.3 Physical responses of sampling sites to mainstem discharge variations Variations in mainstem discharge cause the heads of side channels to alternately be overtopped or dewatered,thereby altering macrohabitat classifications due to changes in water quality,flows,wetted areas, and the amount of cover.The relationships between side channel flows and mainstem discharge at the sampling sites are presented in Quane et a 1.(1985). Changes in wetted area of sites due to variations in mainstem discharge are important because these changes may directly increase or decrease fish habitat.Areas measured from aerial photos have been compiled for selected side channel and slough complexes by Ashton and Klinger-Kingsley (1985)for a variety of discharges.Mainstem backwater effects at tributary mouths are al so important because object cover inundated by backwater is an'important component of these sites for juvenile salmon.Discharge related responses of site area for all sites pooled and cover for selected tributary mouths will be presented in the next two sections. 3.1.3.1 Area The areas of the RJHAB study sites were calculated geometrically at modelled discharges,and then plotted against mainstem discharge by eye. Measurements of area were then read from these graphs in the range 19 CHULlTNA,TALKEETNA PLUME EFFECTS 1.1 JULY 19.1984 LOWER 8U8ITNA Cl WEST BANK +MID-CHANNEL o EAST BANK ~, 0.9 ~0.8 .....-g 07 ~i . Q~0.6 m~~.0.5 0.4 0.3 0.2 0.1 a CHULITNA MIDDLE SUSITNA TALKEETNA - 5 15 2S 3S -CHULITNA o AUGUST 18.1984 1.2 1.1 1.3 -r---,...------------:~=--------___, LOWER 8U8lTNA D WEST BANK +MIO-CHANNEL o EAST BANK 0.9 ~0.8 '-oI~7~.i o. Q~0.6 rtJ~~0.5 0.4- 0.3 0.2 TAU<EETNA MIDDLE SUSITNA 0.1 O+---t----r--.,...--~---r--_r--,_-___,r_-___j 5 15 2S 35 MILES DOWNSTREAM FROt.f CONFLUENCE Figure 4.Comparison of turbidities in the lower Susitna River below the Chulitna and Talkeetna River confluences on July 19 and Auoust 16,1984. 20 - - - - between 12,000 to 75,000 cfs at 3,000 cfs increments.Since Eagles Nest Side Channel was modelled only at discharges less than 20,000 cfs,we did not try to extrapolate values over this range for this site. Similarly,area·response at the six IFIM sites were calculated by the IFG program at side channel flows which corresponded to increments of 3000 cfs within the 12,000 to 75,000 cfs mainstem discharge range. Individual area responses for all the modelling sites have been tabu- lated in Appendix Table B-4 at 3,000 cfs discharge increments.Also, side channel flows associated with these increments have been tabulated. By summing areas of the sites by macrohabitat type,the response of the pooled sites can be illustrated.The combined area of three tributary mouths increased greatly at di scha rges greater than 27,000 cfs (Fi gure 5).Since sloughs transform to side channels at greater discharges, slough habitat decreased with discharge while side channel habitat steadily increased (Figure 6).Slough habitat was broken into two categories:total and access ib1e.The total category i ncl udes ponded water with no access from the mainstem while the accessible sloughs are those with potential access from the mainstem. 3.1.3.2 Cover Since instream cover is an important component of fish habitat,the response of available cover tomainstem discharge at individual sites is of interest.Increases in instream cover (debris,riparian vegetation) at side channels were often accompanied by large increases in flows and related water column velocities.Therefore,increases in suitable cover at side channels were often offset by increases in velocities which made the site unsuitable.Turbid water in side channels also provides cover for juvenile chinook salmon and therefore,instream object cover may be less necessary for chinook salmon under turbid conditions (Suchanek et a 1.1984). At tributary mouths,on the other hand,tributary flows ·are independent of mainstem discharge,the water is often clear,and the primary effect of mainstem discharge is the formation of a backwater zone.Increases in mainstem stage typically decrease velocities and "inundate cover at tributary mouths. Cover responses to mainstem discharge at the four tributary mouths varied.At Birch Creek Slough,there were no changes in cover as a result of changes in mainstem stage during 1984 sampl ing because the sampling site was located high enough (0.7 miles)up the channel that it was not influenced by mainstem stage.At Beaver Dam Slough,increases in total cover caused by rises in mainstem discharge were limited because most of the cover was submerged aquatic vegetation (Figure 7). At Rolly Creek and Caswell Creek mouths,however,the amount of cover increased rapidly at discharges larger than 45,000 cfs.Increases in total cover at Rolly Creek mouth were caused primarily by inundation of emergent vegetation while both emergent vegetation and overhanging riparian vegetation cover became more abundant at Caswell Creek mouth at high mainstem discharges. 21 703050 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (efs) TRIBUTARY MOUTHS (BIRCH SLOUGH EXCLUDED) 300 I _.'"'I 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 ~G a a B ~IIIIII8 10 ?::-....It) "tJ 'Cero14It) -::J ~o~ N N Figure 5.Area within modelled tributary mouths as a function of mainstem discharqe at the USr,S Sunshine gaging station!1984.Boundaries of the site were fixed. 1 J ,1 J t I J J },.~,s J t ~ 30 50 70 (Thousands) MAlNSTEM DISCHARGE AT SUNSHINE (cfs) - ..... - - ..... I SIDE CHANNELS (EAGLES NEST S.C.EXCLUDED)2..,....---------------------=::::>&----, 1.9 1.8 1.7 1.6 1.5 1.4 1.3....... ..1.2-.......&t!1 .1 ~~1 ~0.9 ~0.8 0.7 0.6 0.5 0.4- 0.3 0.2 0.1 O-H:I-l:l-lEF-,.---.,..---.....,.---,....----.--"""""1r----l 10 Fiqure 6.Area within modelled sloughs and side channels as a function of mainstem discharge at the US~S Sunshine qaqinq station~ 1984. 23 - ~ BEAVER DAM SLOUGH COvER RESPONSE ~, 0 .•~ 0.8 0.7 ..<n:JT..... ~0.6 .AQUAT'1,t£C "-III '"0.5 D [wE,R VEe.0: 0:~0.4 <J 0.3 0.2 0.' 0 20 30 40 50 60 70 r.MINSTEM OISC~~~·~UN5HIHE(cts) ~ ROLLY CREEK MOUTH COvER RESPONSE 0."•Tm".... 0.8 0.7 !~0.6 ...0.50: 0:~0.4 0.3 0.2 ,0.1 -40 60 !IO MAlNSTEw ~E::r~NE (cfa) CASWELL CREEK MOUTH CO'w"ER RES-PeNSe: 0.9 0.8 0.7 III Z 0.6 0"-III 0.5'"0: 0: 0.'~<J 0.3 0.2 0' 0 10 •N:JUAT a EMDi •0\<0<_ 30 50 70 MAlNSTE....OISC~~£~·~UNSHII'fE Cd_) Figure 7.Instream cover response at Beaver Dam Slough t Rolly Creek t and Caswell Creek mouths as a function of mainstem discharge at the LJSr,S Sunshine qaging station t 1984. 24 ""'I f -1 ""'" 3.2 Distribution and Abundance of Juvenile Salmon Chinook,coho,chum,and sockeye salmon juveniles were captured at the twenty habitat model sites,But only one pink salmon fry was captured. Pink salmon outmigrate early and our methods are not effective at capturing them.A summary of the juvenile chinook,coho,chum,and sockeye salmon catch and catch per cell (CPUE)data by site is given in Appendix Table B-2. 3.2.1 Chinook salmon Fourteen hundred fifty-ei ght juveni 1e chi nook salmon were coll ected in the lower reach of the Susitna River from June through early October. Approximately 83%of these fish were captured at the 20 habitat model sites.Age 0+fry accounted for 93%of the chi nook salmon j uveni 1es captured.The percentage of 0+fry increased from 66%in late June to 99%in early August.All chinook fry captured after early August were 0+fish,indicating that 1+chinooks had outmigrated from the study reach prior to August 15. Chinook fry were widely distributed at the modelling sites from early June through late August (Figure 8).Last Chance Side Channel was the only site where no chinook juveniles were captured.Chinook juveniles were captured at 80%or more of the sites sampled in early June and late August.In September and early October,the proportion of sites where chinook salmon were captured decreased. Mean juvenile chi nook CPUE was hi ghest at tri butary mouths,where 1.5 fish per cell (fpc)were captured.At side channels,the mean CPUE for juvenile chinook was 0.8 fpc.Slough catch rates were consistently low (0.1 fpc).Mean catch rates at side channels were relatively constant throughout the season,whi 1e tributary mouth CPUE's peaked in August (Figure 9).The peak CPUE for tributary mouths occurred in late August at Caswell Creek mouth (20.2 fpc).The peak CPUE at a modelled side channel (4.4 fpc)occurred at Sunset Side Channel.CPUE I s within the side channels peaked at turbidities of 100 to 150 NTU (Figure 10).The correlation (r)between mean turbidity of the modelled side channels and mean catch per cell of chinook salmon was -0.63 (p <0.05). Catches at Trapper Creek Side Channel appeared to reflect the effect of turbidity upon chinook fry use.This west bank site,located below the Chulitna River,had a high CPUE in early June (2.7 fpc)when turbidity was low but then no chinook were captured in late June and early July when turbidities were above 550 NTU.Chinook fry catches increased slightly on subsequent trips when turbidities began to decrease. 3.2.2 Coho Salmon Four hundred forty-two juveni 1e coho salmon were captured wi thi n the lower Susitna River study areas of which only five were not captured within the habitat model sites.Three age classes of juvenile coho salmon were captured.Eighty-six percent of the juvenile coho captures were age 0+and 14%were age 1+.Only one age 2+juvenile was captured. 25 I t") ~~~•i.""/CHINOOK SALMON ~6\~-.....~\;10\SAMPLING PERIOD...~~.\.'-.:I JUN JUL AUG SEP OCT .SOt t.II I ]I I ]I I :II:I1~~raepper Creek S.C.•.0 D l2][2]{;;ii [;ji [2]0 r1l ~Birch Slough f.ZI ~D 0 I2l ~f.ZI 0 0 ,~.~Sunrise S.C.0 -Ciii CiiiiiI •Ciii - - -1~~'\Sunset s.c.---.•[;;iiI [;iii IZl - ~Beaver Oam SlouGh 0 ~DOD [;iii D 0 - ~Beaver Dam S.C.f2]•CiiiiiI •[,;iiI [,;iiI [,;iiI 10 - ~~Sucker S.C.Gil ~[ZI [;;iiI 0 IZJ I2l - - : Sauna S.C.f7l -D 0 I2l [ZI 0 - - "=t Circular S.C.[;;iiI 0 IZI IZI ~[;iiiI 0 - -~~Goose 2 S.C.[;iii ~[;;iiI lZJ Gi ~0 - - \ ~;;;Mainstem West Bank - - -0 0 IZl [;iii 0 - ~<~r ;I .~I--J Island S.C.0 [;;iiI I:;ii ~~I2l 0 0 - ()'J ,r~Caswell Creek Mouth . • -0 • • •IZI fZI 0 1.'f),~:.tl Rustic Wilderness S.C.[;jjJ [;;iiI •~[;jjJ IZI [;jjJ 0 - ";~~~'~~Last Chance S.C.- -0 10 0 0 - 0 - Q~.~Bear Bait S.C.-0 f2j [][21 D·0 - - ,~...j.j'RolI'l Creek Mouth IZI !ZI 0 IZI [;jjJ [;;iiI IA Gii [ZJ "'iY'Ii Kroto SlouGh Head 0 0 121 0 0 IZI 0 - - .,,"Eagles Nest S.C.- - - - - -0 0 ~ j~~;;;;~:~~u:'6~NC~~EY~0 0 ~~0 0 -0 ~0 0.00 [;iii 0.25-2.50 ~~~[2]0.QI-O.25 •>2.50 V(]-No sample o _3"---- Fiaure 8.Seasonal distribution and relative abundance of juvenile chinook salmon on the lower Susitna River,June through mid-October 1984. 26 - - - ,- ".5 4 ,- :l.::lS:SI TRIBUTARY MOt1l'HS 0 3.50z J: 0 3I ..J..J 2.5I.&J 0 IS 2Q.. J:g 1.5~ ~ ::E 0.5 o E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPUNG PERIOD Figure-9.Juvenile chinook salmon mean catch per cell at side channels and tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984. 75 125 175 225 275 325 375 >400 TURBIDITY (NTU) 1.9 1.8 1.7 1.6 1.5 1.4 ::i 1.3w 0 1.2 f5 1.1 Q.1J: ()0.930.8 ~0.7 0.6::E 0.5 0.4 0.3 0.2 0.1 0 25 - Figure 10.Juvenile chinook salmon mean catch per cell at modelled side channels on the lower Susitna by turbidity increment,June through mid-October 1984. 27 The percentage of age 1+fry captured decreased from approximately 50% in early June to 2%in early October. Juvenile coho salmon were unevenly distributed in the study area,being captured at only 50%of the 20 modelled sites (Figure 11).Only one coho was captured at four of these sites.In'most instances,juvenile coho CPUE's tended to be higher in late summer. Juvenile coho salmon catches varied greatly among the three macrohabitat types.Tributary mouths had a mean juvenile coho CPUE of 1.2 fpc while sloughs and side channels had CPUE's of 0.02 and 0.01 fpc,respectively •. Juvenile coho were captured at all four tributary mouths,five of the 16 side channels (31%)and two of the 14 sloughs (14%)sampled.Over half of the juvenile coho were captured at Caswell Creek mouth,with the majority in mid to late August.The juvenile coho catch rate at tribu- tary mouths ranged from near ten juveniles per cell at Caswell Creek in late August to zero fish per cell at several sites during various sampling periods throughout the open-water season (Figure 12).With the exception of Birch Creek Slough,coho CPUE's were higher during late summer and fall than during early summer sampl ing periods. 3.2.3 Chum salmon Six hundred eight juvenile chum salmon were collected in the lower Susitna River of which only ten were captured at opportunistic sites. In early June,chum fry were captured at 13 of 15 (87%)modelling sites sampled (Figure 13).By late July,chum were only captured at six of 19 (32%)sites sampled.Over 99%of the total catch was made prior to August and no chum salmon fry were captured after August 15.The majority of sites with high CPUE's were located in the reach from Island Side Channel (RM 63.2)to Sucker Side Channel (RM 84.5). Chum fry CPUE's declined steadily from early June to mid-August (Figure 14),reflecting outmigration of juvenile chum salmon from the Susitna system.In a pre-study trip in~late May,chum fry were collected at a number of lower river sites and appeared widely distributed in the river. Juvenile chum CPUE's were highest in side channels (0.6 fpc)and tribu- tary mouths (0.1 fpc).,Slough CPUE's of juvenile chum were low (0.01 fpc),however,sampling effort at sloughs was limited from early June through early July.Tributary mouth densities were unequally distri- buted by a single site catch of 39 fry at Birch Creek Slough in late June.Juvenile chum catches at side channels were affected by turbi- dity.Peak chum catches were made in side channels with a turbidity of less than 50 NTU (Figure 15). 3.2.4 Sockeye salmon Four hundred twelve juvenile sockeye salmon were captured in the lower Susitna River study reach.Ninety percent (369)of these fish were captured at the habitat modelling sites.Age 0+sockeye comprised 99% of the catch.Age 1+sockeye were found in early June at Hooligan Side Channel,a site which produced no further sockeye juveniles all season, 28 - - - - - COHO SALMON ..... I C') ::to /~~e t ;;-~\ol Q ~ ':Q e,r-_....1 -;.".to'SAMPLING PERIOD"',-:/;\"JUN JUL AUG SEP.OCT ~Site I II I 1I I II:1 II:I~K Trapper Creek S.C.0 0 DOD 0 0 0 [ZJ ~Birch Slough •{2]j:;;j 0 [;jjJ 0 [;jjJ [;jjJ r:zJ ,~Sunrise S.C.0 -0 IZl 0 -tj - - - \Sunset S.C.- - -0 0 ODD - CC Beaver Dam Slough 0 ~1Zl'j:;;j I2lGiiil:LZI • - ~Beaver Dom S.C.0 C!IZI 121 0'IZJ eilD - ~,\,;;s:.=.uc.:.;.k:..:.er~S.:,,:::C':""---f!0~.!::D::!.~D~~O~O~.~O~D=~-+-~ Sauna s.c.0 -rl DOD I(J - - Circular S.C.0 0 0 0 0 0 0 - - Goose 2 S.C.0 0 0 12]0,0 0 IZJDDODODO- ---00000- --0000-0- o O'0 0 I;i 1ZI I;i fA [;;iiI D 0 ODD fZI []ID - ------000 ODoocroo-o -000000-- 0000000-- .[;jjJ -~• • • •[;jjJ • Island S.C. Last Chance S.C. Caswell Creek Mouth Bear Belit S.C. Moinstem West Bonk Rustic Wilderness S.C. Hooligan S.C. Rolly Creek Mouth Kroto Slough Head Eagles Nest S.C. MEAN CATCH PER CELL RELATIVE ABUNDANCE KEY 00.00 Gi 0.25-2.50 [2]O.QI -0.25 •>2.50 -No sample (ill - ~, If J) -~''---- Cook InltJf Figure 11.Seasonal distribution and relative abundance of juvenile coho ,-salmon on the lower Susitna River~June through mid-October 1984. 29 CASWELL CREEK MOUTH BEAVER DAM SLOUGH '0 4 g,I J "'.ill .".5 0 8 0 5 g "0 .., I I :l :l 2.5 6'"Q '"Q00 a:5 W a:2 W '"Ie ....IQ.....I J:4 a.li D. 8 ::i 3 1.5 ::i "<< !0 !1 ~0 ::Ii 2 ~::Ii ~b00.5 Z l77'";l Z 0 0 E JUN L JUN E JUl L JUL E AUG L AUG E SEP L SE"E OCT E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPUNG PERIOD SAMPUNG PERIOD BIRCH CREEK SLOUGH ROLLY CREEK MOUTH 4 W I ".5 "5a 0 0 J:"5 "0 0 0 I I :l 2.5 ....I 2.5....I '"'"0 0 a:2 a:2IeIe J:J: il 1.5 5 1.5 is ~1 ! ::Ii ::Ii 0.5 O.S 0 0 E JUN L JUN E JUL L JUL E AUG I.AUG E SEP L SEP E OCT E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAIoIPUNG PERIOD SAMPUNG PERIOD • Figure 12.Juvenile coho salmon mean catch per cell at four tributary mouths on the lower Susitna River by sampling period,June through mid-October 1984. ~3 J J J "J J J J .1 ~I J I f J )J - CHUM SALMON / \~~:::;::e s~~~·(;iI =~fj 8 80 0 =~.Beaver Dam SlouOh 00 0 0 0 10 0 0 - ..,Beaver Dam S.C.f.2][;;iiI 121 0 tzI 0 n In - ~.~\.:S:=.uc::k~e,~s.=c.--:,;.~.~.!!!!!.f:~1Zl~D=:··:+.0;:;;.*0;+:0:=+=--+--=-1 Sauna S.C.• -rei 0 00 0 -_. ·.i~\ I) ":Q e\r--.,.'~l:.~O\SAMPLING PERIOD...,I JUN JUL AUG SEP OCT --II"'.Site I n I n I n I 1I I~~.:..;Tr...:...;o:.....pp-e'-c-r-ee-k-s-.c-.---r::::~=-r:::D=·""D::=-,r:::[;jjjjJ:=-r:D~-=D~D:::=-r;D:;:;;"=D:::-' Birch Slough [;jjjI .0 0 0 0 ClOD ..- - - - - ~Circular S.C.•[;j 12I IZIIOO r--_. .~~Goose 2 S.C.I:i c;ji 0 0 0 0 r--_. ~~p,Ma.instem West Bank - - -10 0 0 0 0 _. ~i ,f1J~'-1_sl_.an_d_S_.C_.-+::::.=+~t:;i fZI (21 0 0 0 - j lIP]Caswell Creek Mouth .0 0 0 0 0 0 0 10~~,)IarJL 'Rustic Wilderness S.C.fA 0 IZl 00 0 0 0 - ~~o:~~~lI~.~::\:;C:.c~·C..=0 ~~8BB-0 0 =- ~.Rolly Creek Mouth IZl 0 DODO 10 0 0";/Ji.Kroto SlouOh Head [21 0 0 0 °°°-- Eaoles Nest S.C.- - - - - -Old 0 Hooli~an s.c.•[;j 0 lZI ClOD -0 MEAN CATCH PER CELL RELATIVE ABUNDANCE KEYo0.00 I:i 0.25-2.50 IZ]om ~0.25 •>2.50 -No sample Cook Inlet Figure 13.Seasonal distribution and relative abundance of juvenile chum .-salmon on the lower Susitna River.June through mid-October 1984. -31 6..,--------------------------. 5 4 3 2 _SIDE CHANNELS lS::sJ TRIBUTARY MOUTHS o - E JUN L JUN E .JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPLING PERIOD Figure 14.Juvenile chum salmon catch per cell at modelled side ~hannels and tributary mouths on thel.ower Susitna River by sampl ing period,June through mid-October,1984. ~, 6..,-------------------------., 125 175 225 275 325 375 400 ruRB10ITY eNTU) 7525 :iw U ~ ct: W 0- J: U'<-u 3 ~ ~ Figure 15.Juvenile chum salmon mean catch oer cell at modelled side channels on the lo~er Susitna River by turbidity increment,June through mid-July 1984. 32 - - .'- - - - - -- and in late June at Beaver Dam Slough.Sockeye juveniles were most widely distributed within modelled sites upstream of Goose 2 Side Channel (Figure 16). Tributary mouths had the greatest densities of juvenile sockeye salmon with a mean catch of 0.7 fpc.The highest CPUE for juvenile sockeye at a tributary mouth was 1.2 fpc at Beaver Dam Slough.Side channels had a mean sockeye CPUE of 0.1 fpc.Beaver Dam Side Channel had the highest CPUE for a side channel of 0.7 fpc.Side slough CPUEs of sockeye juveniles were minimal (0.03 fpc).Side channel CPUEls remained at low levels through August in comparison to tributary mouth CPUEls which varied greatly (Figure 17).No sockeye juveniles were captured in side channels after August,however,sampling was limited. sock.eye fry CPUEs were hi~hest in side channels where turbidities ranged between 100 and 150 NTU (Fi gure 18).The numbers of sockeye juveni 1es captured in.Beaver Dam Side Channel,immediately below and contiguous with Beaver Dam Slough,may have been enhanced by site to site movement. With Beaver Dam Si de Channel captures exc1 uded,the peak CPUE for juvenile sockeye in side channels occurred at turbidities between 50 and 100 NTU. Catches at Beaver Dam Slough and Beaver Dam Si de Channel show the effects of turbidity as cover on the distribution of sockeye juveniles (Figure 19).From late June through August,Beaver Dam Side Channel was breached by the mainstem,the water was turbid,and sockeye CPUEls were high.In early June and September,however,the head of the channel was not breached,the water was clear,and few sockeye juveniles were caught in this environment with little cover.In contrast,Beaver Dam Slough, which had abundant aquatic vegetation cover,had high CPUEls of sockeye juveniles in late August and September.Catches at Rolly Creek also increased in 1ate August and remained fai r1y hi gh through early October (Figure 19). 3.3 Habitat Modelling of Rearing Juvenile Salmon The response of juvenile salmon habitat to variations in mainstem dis- charge was modelled using two techniques:(l)the RJHAB model developed in Marshall et a1.(1984)and (2)the IFIM hydraulic models discussed by Bovee (l982).Suitability criteria for important microhabitat variables are necessary as inputs to both models and criteria specific to the lower reach of the Susitna River for juvenile chinook,coho,chum,and sockeye salmon have been developed in Appendix A. In the following discussion,results are presented by species.Each presentati on inc1 udes modell ing resu1 ts from se1 ected si tes usi ng the RJHAB or IFIM models,pooled results from all the sites modelled,and a test of model verification. No results from the Birch Creek Slough and Eagles Nest Side Channel modelling sites are presented here.At Birch Creek Slough,there was no measurable effect of mainstem discharge upon the site as mainstem backwater at discharges less than 75,000 cfs did not extend to the site and a blocked culvert at the head of the slough stopped mainstem water 33 - - - - /t:-t SOCKEYE SALMON~...J1I-:::l ~\~~o ':l:l>a'-...'<.~~o\SAMPLING PERIOD.....AUG SEP OCT "I .JUN JUL 1~K Site I II I n:I Jr I :II:I Trapper Creek S.C.[2]0 [2]0 0 0 lZ1 0 CI 'i Birch Slough 0 ~0 0 IZ1 I2l 0 0 0 Sunrise S.C.1ZI -IZI IZI IZl [2]-- - \Sunset S.C.---[;iii I2l [2]0 0 - Bedver Dam Slough 0 •[;iii 0 0 (ijjjI Gil [;;jjI -~I~re-Dem s.c.0 fA [;jjJ IA ~~0 0 - ~~.Sucker S.c.[ZJ CjjjI 0 ~10 I2J 0 --~~Y'Sauna S.C.[ZI -0 0 0 [;;jjI 0 -- ~Circu lor S.C.c;j 0 0 0 0 IZJ 0 -- 10.,;[2]0 0 0 0 0 0 --.....Goose 2 S.C. ~Mainstem West Bank -- -0 I[]I2l 0 0 -)~1[J I"sland S.C.IZl 0 0 0 0 0 0 0 - j Caswell Greek Mouth 0 -0 [;iii)[;;iiI [;;iiI 1121 10 0 Rustic Wilderness S.C.0 0 0 0 0 0 0 0 - -='-Last Chance S.C.--0 0 0 0 -0 --='~~Bear Bait S.C.-0 0 0 0 0 0 --~.JZI IZI 10 (:;jjjI 0 [;iii)~[;iI (;iii....U~j Roll y Creek Mouth•-.....Kroto SIoU9h Head 121 0 0 0 0 0 "12]--~.,..Eagles Nest S.C.------0 0 0 ~l HooliQdn S.C.Gil 0 10 0 0 0 0 -0 MEAN CATCH PER CELL RELATIVE ABUNDANCE KEY 00.00 [;iii 0.25-2.50 [2]0.01-0.25 •>2.50 -No sample \ .4• Coolt Inlsf - Figure 16.Seasonal distribution and relative abundance of juvenile sockeye salmon on the lower Susitna River,June through mid-October 1984..~ 34 ,_SIDE CHANNELS IS::sJ TRIBUTARY t.40UTl-lS .3-,------"-.;.--'-''''-''''-''''-'~~--''-'''-'--__. 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 0.8 0.6 a.'" 0.2 a E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT SAMPUNG PERIOD -Figure 17.Juvenile sockeye salmon mean catch per cell at side channels and tributary mouths on the lower Susitna River by samplinq period,June through mid-October 1984. -LESS BEAVER 0N0t s.C cs::sJ WITH BEAVER DIW S.C 75 125 175 225 275 325 375 >400 TURBIDITY (NTU) Figure 18.Juvenile sockeye salmon mean catch per cell at modelled side channels on the lower Susitna River by turbidity increment (with and without Beaver Dam Side Channel),June through niid- October 1984.- 35 - - - .. BEAVER DAM SLOUGH 7 6 0 I 0u 5 I .............40..... G.. I J ~Q~w !2 ... G.. ~::IE < CO.. 0 Z 0 E .IUN L .IUN E .lUL L JUL E AUG L AUG E SEP L SEP E OCT SANPI.JNC PERIOD BEA-";:R OAM SIDE CHANNEL 7 6 0 I 0 0 5 I -'-'...40.. ""a.. I J 5 Qw ~2'... G.. ::E::IE <CO.. 0 Z 0 E .IUN L ./UN E .lUL L .IUL E AUG L AUG E SEP L SEP E OCT SAMPlING PERIOD ROLLY CREEl<MOUTH 7 6 0 I00 5 I...........40.. ""II. I 3 3 ~2 ::IE 0 e:.IUN L .IUN e:.lUL L .lUL E AUG L AUG E SEP L SEP E OCT SAMPUNG PERIOD FiQure 19.Juvenile sockeye salmon mean catch per cell at Beaver Dam Slouqh,Beaver Dam Side Channel,and Rolly"Creek Mouth by sampling period,June through mid- October 1984. 36 - - .i""" - -- ..... - ,.... from flowing through the site.The Eagles Nest Side Channel site was modelled only twice at mainstem flows of 14,900 and 20,400 cfs and therefore could not be readily extrapolated to discharges of 75,000 cfs. All of the other sites were modelled at three or more discharges and results were extrapolated to discharges ranging from 12,000 to 75,000 cfs.The WUAs and site areas at the RJHAB sites were not adjusted to a reach length of 1,000 ft as were the IFIM WUAs.Lengths of all the RJHAB sites are listed in Appendix Table B-3,so that the WUAs could be adjusted if desired. The instream flow results have been generated only to discharges of. 75,000 cfs because it is very difficult to collect data at discharges greater than 75,000 cfs.At 75,000 cfs,most of the side channel sites have very large flows and are poor habitat for juvenile fish.At higher discharges,the entire flood plain becomes full and the flows are barely constrained within the side channels.Refuge for the juvenile fish at these times presumable include large backwater areas and small side channels which are infrequently flooded. At Island and Trapper Creek side channels,both RJHAB and IFIM models were run on the same transects.Comparative resul ts for these two models are given in Appendix C.The summary figures presented here incorporate data from the RJHAB model at these two side channels. The abil ity of the RJHAB model s to extrapol ate WUA between di scharges of 12,000 and 75,000 cfs was rated unacceptable ~o good (Table 5).Some model s were rated fair because there were no habitat measurements taken at discharges just above overtopping of the side channel.Eagle's Nest Side Channel was rated unacceptable because measurements were taken on only two occasions at discharges less than 21,000 cfs. The IFIM models were evaluated according to hydraulic criteria on the basis of excellent to acceptable (Appendix D).Acceptable ranges of the models usually extend to over 60,000 cfs (Table 6).The models were run and WUAs generated at side channel flows which corresponded to dis- charges ranging to 75,000 cfs,so reliability at these flows is unknown. At discharges below overtopping,the WUAs of IFIM sites at flows of 5 or 6 cfs were used,except at Trapper Creek Side Channel where a site flow- mainstem discharge rating curve for unbreached conditions developed by Quane et al.(1985)was used to estimate unbreached flows. Since suitability criteria for chinook salmon juveniles have been developed for both turbid (>30 NTU)and clear «30 NTU)conditions, several assumptions were made.Tributary mouth sites were assumed to be clear (>30 NTU)at all discharges less than 75,000 cfs.This is not always the case,as occasionally turbid mainstem water may back up into tributary mouths with a rapid increase in mainstem stage.Also spring runoff or large storms may increase turbidities at tributary mouths to over 30 NTU.Available data,however,have indicated turbidities at tributary mouths are normally less than 30 NTU (Figure 3).At side channel/slough sites,turbidities were assumed to be greater than 30 NTU when the site was breached and less than 30 NTU when the site was not breached.In early June,September,and early October,turbidities in side channels were sometimes less than 30 NTU (Figure 3).Many of the 37 Table 5.Evaluation of RJHAB model quality for extrapolating WUAs over the range of 12~OOO to 75,000 cfs as measured at Sunshine gaging station,1984. Site Hooligan Side Channel Eagle's Nest Side Channel Kroto Slough Head Rolly Creek Mouth Bear Bait Side Channel Last Chance Si de Channel Rustic Wilderness Side Channel Caswell Creek Mouth Island Side Channel Coose 2 Side Channel Sucker Side Channel Beaver Dam Slough Beaver Dam Side Channel Sunrise Side Channel Birch Creek Slough Trapper Creek Side Channel Number of Habitat Measurements 5 2 4 4 4 5 5 3 5 4 4 4 3 4 3 4 Model Quality' Cood Unacceptable Fai r Cood Fai r Fair Cood Fair Cood Fai r Cood Cood Cood Fair Good Good .... ~, Model quality definitions: 1.Good -Side Channels:Measurements spaced so as to cover the range of mainstem discharges above breaching to 75,000 cfs.Models include information about unbreached,barely breached,and a minimum of two other breached flows,one near 75,000 cfs. Tributary Mouths:Models include information when no backwater,moderate backwater,and high backwater present. 2.Fair -Side Channels:Model missing information concerning habitat when channel is barely breached~or other flows given above. Tributary Mouths:Not enough measurements to accurately describe amount of backwater effect. 3.Unnacceptable -Less than three data points -cannot describe a curve. - - - - Table 6.Discharge ranges of IFIM models hydraulics are rated acceptable,1984. at lowe r Sus itna Ri ve r Data taken from Appendix D. sites for Site Island Side Channel Mainstem West Bank Circular Side Channel Sauna Side Channel Sunset Side Channel Trapper Creek Side Channel 38 Acceptable Range 35,000 to 70,000 cfs 18,000 to 48,000 cfs 36,000 to 63,000 cfs 44,000 to 63,000 cfs 32,000 to 67,000 cfs 20,000 to 66~000 cfs - - - .-, - model sites were not breached during these periods of low mainstem discharge ..Turbidities in side sloughs were usually less than 10 NTU. 3.3.1 Chinook Salmon Chinook salmon juveniles were captured at all of the study sites with the exception of Last Chance Side Channel (Figure 8).Since chinook juveniles were widely distributed,results from all sites modelled with RJHAB and IFIM techniques will be presented. Graphs of the weighted usable'area responses to mainstem discharges for all sites not presented here are included in Appendix B.Appendix B also contains the tabulated values of weighted usable areas at 3,000 cfs increments as digitized from these graphs (including site graphs pre- sented here).Also tabulated are habitat indices which were calculated by dividing the weighted usable area at a given discharge by the site area at the same discharge. At the Rolly Creek,Caswell Creek,and Beaver Dam Slough tributary mouth sites,the responses of weighted usable area to mainstem discharge were very similar.The Rolly Creek mouth weighted usable area response to discharge is presented here as an example (Figure 20).The great increase in weighted usable area with discharge above approximately 45,000 cfs is due to the effect of mainstem backwater causing large increases in area,depth,and amount of cover. At side channel/slough sites,the responses of weighted usable areas to mainstem discharge was varied.Normally,the weighted usable area increased greatly after overtoppi ng and then decreased with further increases in mainstem discharge as at Kroto Slough Head (Figure 20). The increase in weighted usable area after overtopping is due to in- creases in area and also increases in cover suitability as turbidity improves cover.As discharge increases with site flow,velocities initially become more suitable,but then as flows continue to rise, velocities become unsuitable and WUA decreases. At Sucker Side Channel,backwater effects buffer the velocities from becoming too high and so weighted usable area increases after overtop- ping and then remains nearly the same to a discharge of 45,000 cfs after which it rapidly increases (Figure 20).At approximately 60,000 cfs, WUA's begin to decline at this site,however,as velocities and depths become unsuitable.At other sites,WUA held quite constant after overtopping or slowly increased (see Appendix B). When WUA's from three tributary mouths are pooled there is no large change in WUA until approximately 45,000 cfs when the WUA increases greatly with discharge (Figure 21).By dividing the WUA at 3,000 cfs increments by pooled area for the three sites and plotting the habitat index,it becomes apparent that the change in WUA is not simply due to increases in site area.Increases in habitat indices are due to in- creases in the amount of instream cover,more suitable velocities,and deeper water which may also provide cover. 39 ROLLY CREEl<~OUTH40 35 """ ::30,;. ~ ~25 ~-ac ~~20~~£.15Cl I!!~ I: 8 10 ~ 5 -0 10 30 50 70~USands§.......NSTE""DISC GE AT UNSHINE (cfs) ~ KROTO SLOUGH HEAD9 8 -. ::7 a ~6 ~ ~-a 5 -c ~~ ~a 4:£ a 3~ (> ~2 "-0 10 30 50 70 90~USOnds§MAlNSTEM OISC GE AT UNSH/NE (cfs) ~ SUCKER SlOE CHANNEL ~4:< ! ~:3H ,, c ",-~ll , ~a:£2 Cl I!! I:-Cl~ 0 ..... 10 30 50 70 MAlNSTEM OISC~:;~a~s~UNSHINE(cfs) Figure 20.Weighted usable area for juvenile chinook salmon at Rolly Creek ~1outh, Kroto Slough Head,and Sucker Side Channel study sites as a function of mainstem discharge,1984. 40 ..... 0.16 0.15 0.14 0.1.3 X 0.12 w 0 ~0.11 !<0.1~0.09 0.08 0.07 0.06 0.05 10 .30 50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (efa) Fi gure 21.Weighted usable area and habitat indices for juvenile chinook salmon at tributary mouth sites as a function of mainstem discharge,1984. 41 When WUA's from the modelled side channels/sloughs are pooled,WUA's increase greatly to approximately 40,000 cfs and then very gradually decline (Figure 22).Habitat indices for the pooled side channels show a similar rise to a peak at 40,000 cfs but then a rapid decrease to approximately 60,000 cfs when the habitat index levels off.The rela- tively more rapid decrease in the habitat index is due primarily to velocities and depths becoming very unsuitable at the higher dis~harges. Turbidity has been shown to be an important determinant of juvenile chinook distribution (Figure 10).Turbidity varies in the Susitna River from the east bank to the west bank downstream from the Chul itna and Talkeetna river confluences (Figure 4).In formulating the pooled side channel/slough response of juvenile salmon habitat,it was desirable to weight turbidity as it varies from site to site. Although turbidity data for the model sites are limited,an average turbidity for the side channels modelled during the period from June through August was calculated in Appendix Table B-1.A preliminary suitability index for high turbidity was then fit to the data in Figure 10 (Table 7).This index is specific only to the turbidity regimes of lower river side channels and is undefined for application to tur- bidities of less than approximately 100 NTU.When the turbidity indices and mean turbi diti es were combi ned,WUA estimates for the sites were weighted differently (Table 8). When the WUA estimates for each site are adjusted by these factors and the WUA's are again totalled,the WUA and habitat index response ad- justed for turbidity for the side channels combined can again be ex- amined (Figure 23).There is very little change from the previous unadjusted graph in the shape of the WUA response curve,but the magni- tude was reduced by almost 40%.Similarly,the shape of the habitat index responses curve has also been changed very little by these adjustments.The lack of change in shape of these curves suggests that the responses of the side channel WUAs and habitat indices are similar for most of the sites. The mean seasonal chinook salmon habitat index for the 15 side channels and four tributary mouths were cal cul ated and compared wi th mean chi nook catch (Figure 24).The positive relationship was statistically signifi- cant (p <0.001)but not very strong.Most of the correlation was due to the large catch (5.16 fpc)and habitat index (0.19)at Caswell Creek mouth.Another outlier is Beaver Dam Slough with a habitat index of 0.17 and a mean catch of 0.17 chinook per cell. 3.3.2 Coho Salmon Si nce coho salmon were captured in number (more than 20)only at the tributary mouth sites,only results from these sites will be presented here.In Appendix B,values of WUA's and habitat indices at 3,000 cfs increments for these areas are presented. The response of WUA to mainstem discharge at the three tributary mouths varied (Figure 25).At Caswell Creek mouth,WUA rose with discharge due to increases in area and the amount of preferred cover.At Rolly Creek 42 - - - - - ...., SIDE CHANNELS /SLOUGHS CHINOOK SALMON70 60 ..........-,-IT 50~ LS....... ~.tJ 40 l: ~51 ~5 30::Jt. 0~20::t:: Q lJJ:t 10 0 r-10 30 50 70~hQUSCndS§ MAINSTEM DISC ARGE AT UNSHINE (cfs) -0.06 --0.05 0.04 Xw 0 ?; ~0.03 r-t::: (D ~ 0.02 0.01 0 10 30 50 70~USOndS~MAINSTEM DISC GE AT UNSHINE (cfs),- -Figure 22.~/eighted usable area and habitat indices for juvenile chinook salmon at side channel/slough study sites as a function of mainstem discharge,1984. -43 Table 7.Preliminary juvenile chinook salmon turbidity criteria derived from lower Susitna River side channel distribution data for turbidities greater than 100 NTU.These criteria are only applicable to lower Susitna River side channels. - .- Mean Turbidity (NTU) 101 -200* 201 -250 251 -300 301 -350 350 Suitabil ity 1.00 0.65 0.55 0.40 0.15 - *Suitability index for turbidities of less than 101 NTU is undefined and may be greater than 1.0. Table 8.Weighting factors for turbidity by side channel site for analysis of juvenile chinook salmon habitat use,1984. - Mean Site Turbidity (NTU) Hooligan Side Channel 377 Kroto Slough Head 388 Bear Bait Side Channel 254 Last Chance Side Channel 365 Rustic Wilderness Side Channel 118 Island Side Channel 215 Mainstem West Bank 279 Goose 2 Side Channel 194 Circular Side Channel 241 Sauna Side Channel 266 Sucker Side Channel 140 Beaver Dam Side Channel 139 Sunset Side Channel 152 Sunrise Side Channel 121 Trapper Creek Side Channel 499 44 Turbi di ty Weighting Factor 0.15 0.15 0.55 0.15 1.00 0.65 0.55 1.00 0.65 0.55 1.00 1.00 1.00 1.00 0.15 ...., - - 40 35 .........-30 cT II'-' ~25 ~-8I:: ~g 20~S~15c ~CI 10~ 5 0 10 0.04 0.035 0.03 x 0.025 I.&Jc ~ !<0.02 !:: III~0.015 0.01 0.005 0 10 SIDE CHANNELS /SLOUGHS ADJUSTED CHINOOK WUA 30 50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (cfs) 30 50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (cfs) - Figure 23.Turbidity adjusted weighted usable area and habitat indices for juvenile chinook salmon at side channel/slough study sites as a function of mainstem discharge,1984. 45 CHINOOK MODEL VERIFICATION (SIDE CHANNELS AND TRIBUTARY MOUTHS) 6 i I 54 Y =0.15 +15.04x p <0.001 ::J 4 I 2 w r =0.39 0 a::wa.. J:.35 ~2 ~ -I=':> O'l I 1 o 18 ....I T'iii iii Iii iii i I I I o 0.02 0.04 0'.06 0.08 0.1 0.12.0.14 0.16 0.18 SEASONAL MEAN HABITAT INDEX Fiqure 24.Juvenile chinook salmon mean catch per cell versus seasonal mean habitat indices at side channel and tributary mouth modelling sites on the lower Susitna River,1984. .J ,J J ,J }J J J ,J ,J .)J ~) Figure 25.Weighted usable area for juvenile coho salmon at the Caswell Creek,Rolly Creek,and Beaver Dam Slough tributary study sites as a function of mainstem discharge,1984. 47 mouth,the WUA first decreased with discharge due to the formation of zero velocity backwater from a free flowing state without major changes in cover or area.At higher discharges,the WUA increases due to a rise in area and usable cover.At Beaver Dam Slough,these effects of backwater formation and increases in cover inundated offset one another so that there was little change in WUA with discharge. When the WUA's from all three sites are summed (Figure 26),there is little change in WUA until approximately 50,000 cfs when the WUA begins to increase greatly with discharge.When the effect of change in area is taken out by calculating a habitat index,site quality decreases initially as the backwater is formed and then begins to increase as cover is inundated by backwater. The mean habitat index for the season (May 15 to October 15)was cal- culated for the four tributary mouths.Since Birch Creek Slough was a natal.area,only catches from mid-July through mid-October were used in calculating the mean site catch.The mean catch per cell of coho juveniles increased with the mean habitat index but a linear regression was not statistically significant at the 0.05 level (Figure 27).None of the side channels had mean seasonal habitat indices greater than 0.05 and most were 0.03 or less,primarily due to the lack of suitable cover types. 3.3.3 Chum Salmon Chum salmon were widely distributed at all of the side channel sites sampled from early June through July 15 (Figure 13).Therefore,graphs of the WUA response as a functi on of ma i nstem di scha rge for a11 the side channel/slough sites not presented here are included in Appendix B. Also tabulated in Appendix B are values of WUA's and habitat indices at 3,000 cfs increments as digitized from the graphs. Responses of WUA's at the sites to increases in mainstem discharge were variable.At Rustic Wilderness Side Channel,WUA greatly increased after overtopping and then declined with further increases in discharge as velocities and,depths became unsuitable (Figure 28).At other sites, for example Last Chance Side Channel,the increase in WUA after overtop- ping was considerably less while at Trapper Creek Side Channel (Figure 29),.WUA's decreased after overtopping.At Sunset Side Channel,WUA increased after overtopping until about 53,000 cfs when WUA·quickly declined.The other sites a.lso showed variations of these response curves (see Appendix B figures). When WUA's from all modelled side channel/slough sites are pooled,the peak in WUA's for the sites occurs at a discharge of 40,000 to 52,000 cfs (Figure 30).Above this discharge range,WUA's decrease rapidly due to unsuitable velocities and depths.Habitat indices for the same pool ed sites are constant through about 24,000 cfs and then decrease steadily. Chum salmon use of side channels was affected by turbidity (Figure 15), and since turbidity varied from site to site,WUA's for each site were adjusted for turbidity.Since chum salmon outmigration is mostly 48 """ - ~ TRIBUTARY MOUTHS 20 (BIRCH SLOUGH EXCLUDED) ~ 19 "18.J--0-17.,....... h 16 fr~c ~g 15~~~14 a~13:r () !i 12 11 10 ~10 30 50 70~USOncls~tAAlNSTEM DISC GE AT UNSHINE (cts) 0.1 0.09 ~ 0.08 0.07 P"'''xw.0.06a ?; !<0.05 !::::: lD 0.04~ 0.03 0.02 0.01 0 10 30 50 70~usand.~MAtNSTEM DISC GE AT .NSHINE (eta) Figure 26.Weighted usable area and habitat indices for juvenile coho salmon at tributary mouth sites (excluding Birch Creek Slough)as a function of mainstem discharge,1984. 49 COHO MODEL VERIFICATION (TRIBUTARY MOUTHS ONLY) 5 I I 4 :i I CASWELLwcU It:3 -IJ.J Q.. I U 3 2-g :::!. <..n C) 1 4 BEAVER DAM a BIRCH ROLLY a a o +I !!i !!iii iii iii o 0.02 0.04·0.06 0.08 0.1 0.12 0.14 HABITAT INDEX Figure 27.Juvenile coho salmon mean catch per cell versus seasonal mean habitat indices at tributary mouth modelling sites on the lower Susitna River, 1984. J J J ~J I .J !J j )J J 1 )J J ) r'" CHUM WUA RUSTIC WILDERNESS SIDE CHANNEL34 32 30 ......28::::26cT..·24...... i5.-..22 ~.g 20c ~~18~~16 ~IIN.chad 0 14 t~12 CI 10iii~8 6 4 2--10 .30 50 70 »>~hOUSOndS~ MAINSn:t.4 DISC ARGE AT UNSHINE (ets) - LAST CHANCE SIDE CHANNEL22 20 ......18 :::: cT 16 Bra.ched,,-411......ti5.-..14- ~-8 12c ~g ~5 10 :£ 0 8 ~6CI iii~4 2 ~-- 0 10 .30 50 70 90~usandS~MAINSn:t.4 DISC GE AT UNSHINE (ets) - Figure 28 .Weighted usable area for Juvenile chum>salmon at Rustic Wilderness and Last Chance Side Channel study sites as a function of mainstem discharge,1984. 51 CHUM WUA TRAPPER CREEK SIDE CHANNEL 50 ---045 ,...t...40-a- 01....,35 ~~-ll 30c ~Hl ~~25 ::3t:. 0 20~ J: 0 iij 15:t 10 5 10 .30'59 70 SUNSET SIDE CHANNEL 54 52 50 -,... 48...-a-46 01...., ~44 42c ~g 40ID:I~~38 ~ ::3t:. 0 36 ~34J: 0 iij 32:t 30 28 26 10 .30 50 70 ~hOUSOndS~ MAINSTEt.l DISC ARGE AT UNSHINE (cfs) Figure 29.Weighted usable area for juvenile chum sallmn at the Trapper Creek and Sunset Side Channel study sites as a function of mainstem discharge,1984. 52 - 420 410 400 ......390::: r:T 380 n '-'370 i:S,...360~-8c:350~~ ~~340 :£330 CI 320~r 310C) W 3:300 290 280 270 10 0.7 0.6 0.5 Xw CI ~ ~0.4- !::::: lD :f 0.3 0.2 0.1 10 SIDE CHANNELS /SLOUGHS CHUM SALMON 30 50 70 (Thousonds) MAINSlEM OISCHARGE AT SUNSHINE (c:fs) 30 50 70 (Thousonds) MAINSlEM DISCHARGE AT SUNSHINE (cts) Figure 30.Weighted usable area and habitat indices for juvenile chum salmon at side channel/slough study sites as a function of mainstem discharqe,1984. 53 completed by July 15,turbidity data contained in Appendix Table B-1 were examined through July 15.Since turbidities greater than 200 NTU appear to affect use greatly (Figure 15),site WUA's were adjusted for periods when the turbidity exceeded 200 NTU.Adjustment factors for the sites ranged from 0.50 to 1.0 (Table 9). When the chum salmon WUA 's were adjusted for turbidity and again to- talled,very few changes were noted in the shape of the WUA of habitat index response curves although both WUA'S and habitat indices decreased (Figure 31).Since there was little change in these curves,it appears that the shapes of the chum WUA responses at all the side channels are very similar and therefore weighting the sites differently by turbidity only changes the magnitude of the response. Mean chum salmon adjusted habitat indices were calculated for the period from May 15 through july 15 and compared with mean chum catch during the same time period (Figure 32).There was no sampling effort at two of the side channels,Mainstem West Bank and Sunset Side Channel,during this time so they are not included in this graph.The correlation (0.54)between the seasonal habitat index and chum catch was significant at the 10%probability level but not at the 5%probability level. 3.3.4 Sockeye Salmon Sockeye salmon were most numerous at the tributary mouth sites with most side channels having some use (Figure 16).Presented here or in Appen- di x B are graphs of the WUA responses to di scharge of the three tribu- tary mouths and the four side channels (Beaver Dam,Sucker,Sunrise and Sunset)whi ch were found to have sockeye salmon present more than half the times sampled. The typical response of WUA at the tributary mouths to increases in discharge was a steady increase as shown here by the modelling results from Rolly Creek (Figure 33).The WUA increased as the backwater zone increased because sockeye find zero velocity water most suitable and because site area and cover also increased greatly with discharge.The WUA response at Sucker Side Channel was similar to that of the tributary mouths as WUA generally increased wi th di scharge after overtopping. Th iss i te is i nfl uenced grea t 1y by bac kwa ter effects from the side channel at its mouth.At Beaver Dam Side-Channel,WUA increased after overtopp"ing and then declined somewhat (Figure 34).At Sunset Side Channel,WUA fluctuated irregularly with discharge as the small amount of usable habitat along the margins of the site moved back and forth with flow changes. At the combined tributary mouth sites,both WUA and habitat indices increased above discharges of approximately 30,000 cfs (Figure 35).At the pooled side channel/sloughs,on the other hand,WUA's also increased after approximately 30,000 cfs while habitat indices generally declined from the peak at 12,000 to 24,000 cfs (Figure 36).The decrease in the habitat index is due to the steadily increasing velocities in the side channels with increases in flow.No adjustments in turbidity are necessary for the four side channel/slough sites as these have very 54 ~l Table 9.Weighting factors for turbidity by site for analysis of juvenile chum salmon habitat use,1984. Site Hooligan Side Channel Kroto Slough Head Bear Bait Side Channel Last Chance Side Channel Rustic Wilderness Side Channel Island Side Channel Mainstem West Bank Goose 2 Side Channel Circular Side Channel Sauna Side Channel Sucker Side Channel Beaver Dam Side Channel Sunset Side Channel Sunrise Side Channel Trapper Creek Side Channel 55 Sampling Period When Turbidity Exceeds 200 NTU June 16-30 June 16-30 June 16-30 June 16-30 July 16-30 July 1-15 June 16-30 July 1-15 July 1-15 June 16-30 July 1-15 July 1-15 July 1-15 July 1-15 June 16-30 Turbidity Weighting Factor 0.50 0.50 0.50 0.50 1.00 0.75 0.50 0.75 0.75 0.50 0.75 0.75 0.75 0.75 0.50 Figure 31.Turbidity adjusted weighted usable area and habitat indices for juvenile chum salmon at side channel/slough study sites as a function of mainstem discharge~1984. 56 }----}l 1 J 'I -t [] o o o I 0.34 a o .. 0.3•n I I ,J 0.14 0.18 0.22 0.26 SEASONAL MEAN HABITAT INDEX J o CHUM MODEL VERIFICATION (SIDE CHANNELS/SLOUGHS ONLY) -r 0.1 4.5 4 - 3.5 - ::fw 3-() ffia..2.5 - J: I a () 3 2 - 01 ~1.5 - "I ~ 1 - 0.5 -,a o I o 0 1 0.06 Figure 32.Juvenile chum salmon mean catch per cell versus seasonal mean habitat indices at side channel and slouqh modellinQ sites on the lower Susitna River,1984.. SOCKEYE WUA ROLLY CREEK MOUTH 110 100 "90 ..;-0-ao II..... ~70 ~.g 60c: ~g ~~50 :€ 0 40 ~.30~ ILl~20 10 --------'l 0 10 .30 50 70 ~usancls~MAtNSTEM OISC GE AT UNSHINE (c:fs) SUCKER SIDE CHANNEL 6 5.........;-0- Il.....4 ~ ~.gc: ~5l :3~~:£Sre.cud 0 2 t~ "G:i~ ---- ~, .30 50 70 (Thousands) IAAINSTEf.4 DISCHARGE AT SUNSHINE (cts) O+----,.--""'-.,....----r---r----r---.,r-----! 10 Figure 33.Weighted usable area for juvenile sockeye salmon at Rolly Creek '·1outh and Sucker Side Channel study sites as a function of mainstem discharge,1984. 58 ~I SOCKEYE WUA BEAVER DAM SIDE CHANNEL ,.... 4.5 4 ,........3.5-0- Il.....3 --------------- ii-.~-8 2.5l: ~g ~g 2:liS 0 1.5~ J: Cl i 0.5 Breached .:5()50 70 (Thousands) MAlNSTEM DISCHARGE AT SUNSHINE (c15) O-t-----,.---,..----r--....,;,.,..r---.....----,,----! 10 SUNSET SlOe:CHANNEL10.-------------------------, ,....9....-0- ,.!!,. ii-.8~~l: ~g ~5=£.7 0~J: !""""Qw~6 .30'50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (cfs) 5+-----,.---T-:----,---,-r----r-----,----j 10 Figure 34.Weighted usable area for juvenile sockeye salmon at the Beaver Dam and Sunset Side Channel study sites as a function of mainstem discharge,1984. 59 ~ TRIBUTARY MOUTHS 130 (BIRCH SLOUGH EXCLUDED)-120 110.........-100 cT ~90 ~~~80 -c ~g 70~5 60:£ c 50~ (.)40~~ 30 20 10 10 30 50 70~USandS~MAtNSTEM DISC GE AT UNSHINE (cfs) ~ 0.42 0.4- ~0.38 0.36 0.34- x 0.32....c ~0.3 ~0.28t:: ~0.26 0.24- 0.22 0.2 0.18 ~ 0.16 10 30 50 70 ~usands~MAlNSTEM DISC GE AT UNSHINE (cfs) ~ Figure 35.Weighted usable area and habitat indices for juvenile sockeye salmon at tributary mouth study sites on the lower Susitna River as a function of mainstem discharge.1984. 60 ~, SIDE CHANNELS /SLOUGHS SOCKEYE SALMON 30 50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (efs) 20 19 ""'18~ 0-17II....., ~16~-8c ljg 15~g:£14 a ~13:r 8 3:12 11 10 10 7030'50 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (efs) 0.2 -r-----,-----------------------, 0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 0.11 0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 O-l-----,----..----,----r---~---._--__T--__; 10- Figure 36.Weighted usable area and habitat indices for juvenile sockeye salmon at side channel and slough study sites on the lower Susitna River as a function of mainstem discharge, 1984. 61 similar turbidity regimes,being located on the same general location on the river.Use of many of the other side channels is probably limited by turbi di ty. The mean seasonal habitat index for sockeye salmon at the four tributary mouths and four side channel sites was calculated for the period from May 15 to October 15,1984.The mean catch of sockeye salmon juveniles was positively related to the mean habitat index (Figure 37).High turbidities and velocities within the other side channels presumably limited use by sockeye salmon juveniles. 62 - ..... - ,~ ~" - -.i· $1 1 1 I ))~J j ~l J SOCKEYE MODEL VERIFICATION (SIDE CHANNELS AND TRIB MOUTHS) y =0.10+2.21x P <0.001 2r ;:0.78 o o o o 0.2 SEASONAL MEAN HABITAT INDEX 0.4 ' Figure 37.Juvenile sockeye salmon mean catch per cell versus seasonal mean habitat indices at side channel and tributary mouth modelling sites on the lower' Susitna River,1984. 4.0 DISCUSSION 4.1 Chinook Salmon Chinook salmon were widely distributed throughout tributary mouths and side channels of the lower Susitna River.Densities of juvenile chinook were highest within tributary mouths.This distribution of chinook fry substantiates earlier observations (ADF&G 1981a;Dugan et al.1983)'that densities of chinook are generally highest at tributary mouths.Caswell Creek mouth had the highest CPUE of juvenile chinook salmon in the lower river and'appears to be a major rearing or holding area. Chinook salmon juveniles used side channels for rearing in both the middle and lower Susitna River after moving from tributary natal areas. Redistribution of chinook fry from natal areas to lower density rearing areas has also been observed in the Deshka River (Delaney et ale 1981) and Montana Creek (Ri is and Fre i se 1978).Th is phenomenon refl ects a downstream movement or dispersal of the 0+age fish (ADF&G 1981c).Most of the 1+chinook juveniles have outmigrated by August l. Use of tributary mouths is limited by the amount of instream cover and suitable velocities.Also,depth may be important to chinook juveniles in tributaries because it probably provides cover in slightly turbid water (10 to 20 NTH)(Appendix A).At Caswell Creek mouth,catches of juvenile chinook were low in September when the mainstem water stage dropped and depths decreased,velocities increased,and amount of cover was reduced. Use of Susitna River side channels by chinook juveniles for rearing is widespread although it is limited by turbidity in portions of the lower reach.Side channels located in the Talkeetna River plume had much higher use than those located in the more turbid Chulitna River plume or those located further downstream where the water of these two tri- butaries are mixed.Si de channel catch rates of juvenil e chi nook (i n similar habitat)in the middle Susitna River in 1983 were approximately four times higher than those in the lower river in 1984 (Dugan et al. 1984). Since lower Susitna River side channels are used less by ,chinook juve- niles than middle river side channels,it is not surprising that sloughs are also used less in the lower reach than in the middle reach.As water levels decreased in the fall and side channel heads dewatered, there were very few chinook fry at slough sites in the lower river to take advantage of the lowered turbidity.Also the side sloughs in the lower river contain little cover. Instream flow effects upon juvenile chinook salmon are related to backwater effects at the tributary mouths and side channel/slough sites and to breaching and side channel flows.When a side slough is not overtopped by the mainstem,access is usually poor and cover is limited. 64 -' - - ...... ,~ .- At tributary mouths,backwater effects increase chi nook use si gni fi- cantly because of increases in instream cover and depth and decreases in water velocity.Also,turbid backwater from the mainstem sometimes intrudes into the sites with rapid rises in mainstem stage.Pooled data from three tributary mouths showed major increases inWUA at mainstem discharges greater than 45,000 cfs. If the study sites had been chosen further upstream in the tributary mouths,WUAs would have begun to increase at a higher discharge,so the 45,000 cfs figure is not absolute.At Birch Creek Slough,for example, there were no measurable effects of backwater to mainstem discharges of 72,000 cfs.In general,increases in mainstem discharge increase the amount of juvenile chinook salmon habitat at tributary mouths.Also, these backwaters may increase access into tributaries where rearing could occur by decreasing water velocities at the mouth. Within side channel/slough sites,mainstem discharge is very important. When sloughs are breached,the water becomes turbid and cover for the chinook juveniles is improved.High turbidities,however,may also limit use of side channels (Figure 10).High turbidities generally occur from mid-June through September (especially duri ng hi gh di s- charges),while turbidities are much lower during the rest of the year. Turbidity also varies spatially within the river.Chulitna and Talkeetna river plume effects extend at least 20 miles downriver (Figure 4).Sites located within the Talkeetna River plume have much lower turbidity and higher juveni 1e chi nook salmon use. Mainstem discharge initially increases chinook WUA within a side chan- nellslough after it overtops but with further increases in flow,WUA usually remains constant or .declines while the proportion of usable chinook habitat declines.The RJHAB model shows a decline in WUA with increasing discharge which is greater than that shown by the IFIM model (Appendix C). The results obtained by pooling WUA from all modelled sites should not be directly extrapolated to represent the entire lower reach.If the modelling sites would have been chosen randomly,many more large,high velocity side channels with extremely little usable habitat would have been modelled.This study was designed to sample proportionately more side channels with.usable habitat which would represent a diversity of channel types in the lower river.The modelled side channels represent a wide range of sizes and shapes of channels with diverse breaching flows,and so these resul ts need to be coupl ed wi th a stratification of lower river side channels by breaching discharge and channel size and type.The most important side channel complexes in the lower Susitna River for juvenile chinook salmon rearing are located within the low turbidity plume of the Talkeetna River.Other side channels or side channel complexes should be weighted according to their mean turbidity level. 4.2 Coho Salmon Juvenile coho salmon in the lower river were found mostly within tribu- tary mouths.Tributaries and tributary mouths were al so the most 65 important rearing areas for juvenile coho salmon in the middle Susitna River (Dugan et a1.1984).Upland sloughs were also used by coho salmon for rearing in the middle river t but upland slough habitat is limited in the lower river and was not sampled during this study. The heavy use of tributary mouths by juvenile coho is due in part to coho in tributary mouths rearing near their natal areas.Their limited use of side channels maybe due to their documented tendencies to favor waters with relatively low turbidities.Sigler et a1.(1984)t for example,found that a larger number of juvenile coho salmon emigrated from exper'imenta1 laboratory channels with turbidities of 25-50 NTU than from clear water channels.In another laboratory study,Bisson and Bilby (1982)established that coho salmon avoided turbidities exceeding 70 NTU.Turbidities in lower Susitna River side channels during June through August often greatly exceeded 100 NTU. Use of tributary mouths by juvenile coho varied greatly seasonally and from site to site.Rolly Creek and Beaver Dam Slough CPUEls of coho salmon generally increased from early summer to late fall (Figure 12). This occurrence may be due to both the immigration of coho juveniles and a decrease in site area.The area of Rolly Creek was reduced by approx- imately 63%from late June and July to September and early October t while the area of Beaver Dam Slough was reduced by approximately 33% between these two time periods.In Birch Creek Slough t on the other hand,a relatively high CPUE occurred in early summer with much smaller values throughout the summer and fall.The relatively high CPUEls in early summer at Birch Creek Slough are probably due to a natal effect. Barrett et a1.(l985)reported that Birch Creek has a spawning run of coho salmon. A comparison of juvenile coho catch rates between tributary mouths and the Talkeetna outmigrant trap (RM 103.0)suggests that a redistribution of juvenile coho into suitable rearing habitat peaks from late July to early August.The catch per hour of age 0+coho at the Tal keetna outmigrant trap increased during this time period while CPUEts at tributary mouths also changed greatly.Birch Creek Slough t which habitat modell ing indicates to be rel ative1y poor coho tributary mouth rearing habitat (Figure 27),shows a reduction in CPUE in late July, following natal emigration t while Caswell Creek,a site evaluated as having relatively good rearing habitat,had increasing CPUEls beginning in late July.A study conducted by Delaney and Wadman (1979)in the Little Susitna River found emigration of emergent fry from natal areas after the end of June. Instream flow effects of the lower Susitna River upon juvenile coho sa 1mon are 1imited to the backwater zone effects at tri buta ry mouths because coho juveniles make little use of the side channel/slough sites. Initi ally,backwater may decrease the amount of habitat sl i ght1y as tri butary mouths change from free flowing to a backwater zone but then WUA generally increases with mainstem stage as cover is inundated. Also,the backwater can provide access into small tributaries and beaver ponds where rearing and overwintering can occur. 66 ...., - - f'""' I Studies of coho salmon distribution in 1982 by hydraul ic zone showed that coho generally preferred free-flowing tributaries over backwater zones (ADF&G 1983).Cover in the free-flowing tributaries is often better than in the backwater areas.For example,Birch Creek Slough generally has poor cover while Bi rch Creek itsel f has abundant emergent and aquatic vegetation in which coho were abundant. 4.3 Chum Salmon The use of minnow trapping during 1981 and 1982 juvenile anadromous studies makes comparisons of lower river catch and CPUE data with 1984 studies difficult because chum salmon are rarely captured in minnow traps.The necessity for very early sampling,almost concurrent with ice-out,becomes important when studying chum salmon juveniles.Their early season movement and short time in the Susitna River system makes detailed conclusions difficult. The large catches of chum salmon fry in side channels in the lower river contrast with the 1983 distribution data from the middle reach.Dugan et al.(l984)indicated that chum fry CPUE's were greatest at tribu- tari es and side sloughs.The 1983 catch rates,.however,refl ect the prevalence of natal sloughs in the middle reach,while the lower reach contains few natal side sloughs.Also,side channels in the middle reach were not extensively sampled until July in 1983. In 1984,chum salmon spawning was observed in several side channell slough sites where none had been observed previously (Barrett et ale 1985)indicating that under certain conditions,lower river side chan- nels do provide some suitable spawning habitat.Chum salmon fry observed in some of the side channels may be rearing near their natal areas. The exact stimulus for the outmigration of chum salmon from the Susitna River is not known,but probably reflects a combination of factors (Roth et al.1984).Mainstem discharge was highly positively correlated with chum salmonCPUE at the Talkeetna outmigrant traps in 1983.The sharp decline in CPUE at the lower r5ver sites from early June (3+fpc)to late June (1+fpc)in 1984 followed the peak June discharge on June 17 at Sunshine Station,and the mid-J~ne peak of chum outmigration past the Talkeetna traps. Since juvenile chum salmon outmigration is mostly completed by mid-July, flow effects are limited to spring and early summer for this species. Juvenile chum salmon used side channels heavily during this time while use of the tributary mouths was limited.Apparently,chum salmon do not move into the tributary mouths as they gradually move downstream and out of the system.Most of the us.e of side channels for rearing occurs before high turbidities occur. Use of side channels by juvenile chum salmon is limited by depth and velocity.The presence or lack of instream cover in side channels is 67 not important to juvenil e chum (Appendix A).Chum fry were captured primarily in shallow sampling cells (S 1.0 ft)which had a relatively low velocity and low to moderate cover.After breaching,side channel WUA's may increase or decrease but the proportion of the area that is suitable generally decreases as velocities and depths become unsuitably large.Turbidities show sharp seasonal increases and some side channels become turbid earlier in the season than others depending upon the turbidity regimes in the Chulitna,Talkeetna,and Susitna rivers. Since chum salmon side channel WUA's respond very similarly to those of chinook salmon at individual sites,it appears that an analysis of response to changes in mainstem discharge for chinook would also hold for chum salmon.An analysis of flow regimes,would only need to take place through mid-July for chum salmon,however,while chinook salmon fry occur throughout the season in side channels. 4.4 Sockeye Sa.lmon Tributary mouths were the primary capture sites for sockeye salmon in the lower river.In the middle river,sockeye salmon were captured primari ly at side sloughs (Dugan et a 1.1984)•.Si de sloughs were the primary spawning areas for sockeye salmon in the middle river,and tributary/l ake systems were the major sockeye spawni ng areas in the lower reach (Barrett et al.1985).Relatively large catches of juvenile sockeye in the middle river side sloughs were due to fish rearing in their natal areas. Few sockeye juveniles were captured in early June at modelled JAHS sites.This low incidence was probably due to lack of natal habitat in mainstem influenced areas of the lower river.Outmigrant trap catches at Talkeetna (RM 103.0)and Flathorn (RM 22.4)indicate that sockeye fry were redistributing in the system by the middle of June (Part 1 of this report).The greatest catch per cell of juvenile sockeye occurred at the modelled sites during late June. The consistently low CPUE's in lower river side channels suggest these areas are of l'imited value for juvenile sockeye rearing.Possibly these juvenil e sockeye catches represent trans ient popul ations.Exceptions include Beaver Dam Side Channel and other side channels located in the Talkeetna River plume where lower turbidities allow juveniles to rear. Since turbid glacial lakes are much less productive for sockeye salmon than are clearwater lakes (Lloyd 1985),the productivity of these side channels for sockeye is probably low in comparison to similar clearwater streams. The larger catches (21 to 101)of sockeye at tributary mouths indicate that these sites are of some value for juvenile sockeye rearing.Beaver Dam Slough had moderate numbers of sockeye present throughout much of the season.Beaver Dam Slough resembled a lake system as it had low velocities,large amounts of cover,and relatively warm temperatures during the open-water season.CPUE1s of sockeye fry at Rolly Creek mouth was low until early August.Emergent and aquatic vegetation were profuse at this site during mid-season,making sampling difficult. After late August,CPUE's of sockeye juveniles increased.Although high 68 """'. - - - .-m, numbers of these salmon fry were caught late in the season~we do not know if they overwinter. Instream flow effects upon sockeye salmon rearing occur at both tribu- tary mouths and side channels.Occurrence of sockeye juveniles in side channels appears to be limited by factors such as turbidity and velo- city.Juvenile sockeye were captured more than half the times sampled only in four side channel sites in the Talkeetna River plume.Even at these four sites,the number of sockeye fry captured was less than 20 at each,except at Beaver Dam Side Channel where 71 were captured.Typi- cally,WUAs for sockeye increase after overtopping of the side channels but then gradually decrease with further increases in discharge as side channel velocities became unsuitable.Sometimes backwater areas may form at the mouths of side channels (for example, Sucker Side Channel) and modify this relationship somewhat so that WUA may rise with increases in discharge for much longer periods.Generally,the proportion of area that is usable within side channels decreases with flow as velocities become less suitable. At tributary mouths,the formation of backwater zones has a major effect in increasing WUA for sockeye salmon juveniles.The response of the increase in WUA for sockeye is similar to that of chinook salmon. Access into suitable rearing and overwintering areas may also occur with the increase in backwater or the amount of overtopping.For example, access into potential rearing areas such as Whitsol Lake may be inhibited if Kroto Slough is not overtopped.Also several other small tributaries along the Kroto Slough side channel may be inaccessible if flows are below those required for overtopping. 69 - - ..- 5.0 CONTRIBUTORS Resident and Juvenile Anadromous Fish Project Leader (Acting Project Leader, Jan.to Jun~,1985) Task Leader Data Collection IFIM Data,Models,and Text Stage Data IFIM Site Selection and Transect Placement Data Base Management 70 Dana Schmidt Stephen Hale Paul Suchanek Rich Sundet (Task Leader) Stuart Pechek Dave Sterritt (Crew Leader) John McDonell Karl Kuntz (Crew Leader) Bob Ma rsha 11 Dale Corzine Mike Domeier Andy Hoffmann (Task Leader) Jim Anderson Jeff Bigler Fred Metzler Tim Quane (Task Leader) Pat Morrow (Crew Leader) Isaac Quera1 Tommy Wi th row Glenn Freeman Doug Sonnerholm Sharie Methvin Chris Kent Di ane Hi 11 i ard E.Woody Tri hey Allen S-i ngham (Project Leader) Chuck Miller Alice Freeman Kathrin Zosel Ga il Hei nemann Donna Buchholz Drafting Typing Text Report Coordinators and Editors 71 Carol R.Hepler Roxanne Peterson Skeers Word Processing Paul Suchanek Karl Kuntz John McDonell Stephen HaJe Drew'Crawford - - -, - """' - - f"""" I i - ...... 6.0 ACKNOWLEDGEMENTS Funding for this study was provided by the State of Alaska,Alaska Power Authority. We thank the various consulting agencies working on the Susitna Hydroelectric Project for helpful comments on a draft of this report. 72 - - 7.0 LITERATURE CITED Alabaster,J.S.1972.Suspended solids and fisheries.Proceedings of the Royal Society of London.B.180:395-406. Alaska Department of Fi sh and Game.1981a.Juvenil e anadromous fi sh study'on the Lower Susitna River (November 1980-0ctober 1981). Phase 1 final draft report.Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. ·1981b.Aquatic habitat and instream flow project.Phase 1 ---;:-final draft report.Volume 1·(December 1980-0ctober 1981). Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage~Alaska. •1983a.Resi dent and juvenile anadromous fi sh studi es on the --=Susitna River below Devil Canyon,1982.Susitna Hydro aquatic studies phase II basic data report.Volume 3 (1 of 2).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies. Anchorage,Alaska. •1983b.Resident and juvenile anadromous fish studies on the --""'Susitna River below Devil Canyon,1982.Susitna Hydro aquatic studies phase II basic data report.Volume 3 (2 of 2:Appendices A-H).Al aska Department of Fi sh and Game Susitna Hydro Aquati c Studies.Anchorage,Alaska. •1983c.Synopsis of the 1982 aquatic studies and analysis of --'"lrfish and habitat relationships (2 of 2:Appendices A-K).Susitna Hydro aquatic studies phase II report.Alaska Deparbnent of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. __........1984a.Susitna Hydro aquatic studies procedures manual (May 1983 -June 1984)(l of 2).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. __."...1984b.Susitna Hydro aquatic studies procedures manual (May 1983 -June 1984)(2 of 2:Appendices).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska. •1985.(In Preparation).Susitna aquatic studies procedures --=manual (May 1984 -June 1985).Alaska Department of Fish and Game, Susitna Aquatic Studies.Anchorage,Alaska. Ashton,W.S.,and S.A.Klinger-Kingsley.1985.Response of aquatic habitat surface areas to mainstem discharge in the Yentna River confluence to Talkeetna reach of the Susitna River.Draft report prepared for Alaska Power Authority.Anchorage,Alaska. Barrett,B.M.,F.M.Thompson,and S.N.Wick.1985.1984 salmon escape- ment studies in the Susitna River drainage {Draft}Report No.4. Alaska Department of Fish and Game Susitna Hydro Aquatic Studies. Anchorage,Alaska. 73 1979.Little Susitna River juvenile Alaska Department of Fish and Game. Bisson,P.A.,and R.E.Bilby.1982.Avoidance of suspended sediment by juvenile coho salmon.North American Journal of Fisheries Manage- ment 4:371-374. Bovee,K.D.1982.A guide to stream.habitat analysis using the in- stream flow incremental methodology.Instream Flow Information Paper.No.12.U.S.Fish and Wildlife Service.FWS/035-82/26. Delaney,K.J.,K.Hepler,and K.Roth.1981.Deshka River chinook and coho salmon study.Alaska Department of Fish and Game,Division of Sport Fish.Federal Aid in Fish Restoration,Project AFS-49,Vol. 22. Delaney,K.J.,and R.Wadman. chinook and coho study. Division of Sport Fish. Dugan,L.J.,D.A.Sterritt,and M.E.Stratton.1984.The distribution and relative abundance of juvenile salmon in the Susitna River drainage above the Chulitna River confluence.Part 2 in D.C. Schmidt,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.)--.1984. Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish &Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage,Alaska. Klinger,S.,and E.W.Trihey.1984.Response of aquatic habitat surface areas to mainstem discharge in the Talkeetna to Devil Canyon reach of the Susitna River,Alaska.E.Woody Trihey & Associates.Anchorage,Alaska. Lloyd,D.S.1985.Turbidity in freshwater habitats of Alaska.A review of pub 1 i shed and unpub 1i shed 1 i terature relevant to the use of turbidity as a water quality standard.Alaska Department of Fish and Game Habitat Division Report No.85-1.Juneau,Alaska. Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L. Crawford,and P.M.Suchanek (eds:T.1984.Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna·Hydro Aquatic Studies..Report No.2.Anchorage,Alaska. Milhous,R.T.,D.L.Wegner,and T.Waddle.1981.User's guide to the physical habitat simulation system.United States Fish and Wild- life Service.Biological Services Program FWS/OBS-81/43. Mundie,J.H.1969.Ecological implications of the diet of juvenile coho in streams.p.135-152.In T.G.Northcote (ed.),symposium on salmon and trout in streams.H.R.MacMillan Lectures in Fisheries,Univ.B.C.,Vancouver. 74 - ~. - - "... - Noggle,C.C.1978.Behavioral,physiological and lethal effects of suspended sediment on juvenile salmonids.Master's thesis,Univer- sity of Washington,Seattle,Washington,USA. Quane,T.,P.Morrow,1.Queral,T.Keklak,and I.Withrow.1985. Technical memorandum in support of Task 14 (Lower River Resident and Juvenile Anadromous Fish Studies).Alaska Department of Fish and Game Susitna Aquatic Studies.Anchorage,Alaska. Riis,J.C.,and N.V.Friese.1978.Preliminary environmental assessment of hydroelectric development on the Susitna River. Alaska Department of Fish and Game.Div.of Sport Fish and Comm. Fish.,Anchorage,Alaska. Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of juvenile salmon from the Susitna River above the Chul itna River confluence.Part 1 in D.C.Schmidt,S.S.Hale,D.L.Crawford and P.M.Suchanek (eds.).-1984.Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage, Alaska.. Schmidt,D.C q S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.). 1984.Resident and juvenile anadromous fish investigations (May - October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage,Alaska. Sigler,J.W.,I.C.Bjornn,and F.H.Everest.1984.Effects of chronic turbidity on density and growth of steelheads and coho salmon. Transactions of the American Fisheries Society.113:142-150. Suchanek,P.M.,R.P.Marshall,S.S.Hale,and D.C.Schmidt.1984. Juvenile salmon rearing suitability criteria.Part 3 in D.C. Schmidt,S.S.Hale,D.L.Crawford and P.M.Suchanek (eds.)-.1984. Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Prepared for the Alaska Power Authority. Anchorage,Alaska. 75 - - APPENDIX A LOWER SUSITNA RIVER JUVENILE SALMON REARING SUITABILITY CRITERIA """" - - INTRODUCTION Habitat suitability criteria are necessary for evaluating fish habitat using the instream flow incremental methodology (Bovee 1982).The criteria express the value of a habitat variable such as velocity on a zero (unusable)to one (optimum)basis for a given fish species and life stage.The suitability criteria are coupled with the habitat present within a study site to produce estimates of equivalent optimal habitat called weighted usable area (WUA). Juvenile salmon rearing suitability criteria have been used to model the response of juvenile salmon habitat to variations in mainstem discharge of the middle reach (Chulitna River confluence to Devil Canyon)of the Susitna River (Hale et al.1984,Marshall et al.1984).The suitability criteria used in these studies were developed specifically for the middle Susitna River by Suchanek et al.(1984).EWT&A (1985)modified a few of the same suitability criteria for use in impact analysis of chinook salmon rearing in the middle Susitna River. In 1984,some of the juven"ile salmon habitat modeling effort was direct- ed toward evaluating responses of juvenile salmon habitat in the lower Susitna River (below the Chulitna River confluence)to discharge varia- tions.Since habitat data collection techniques used in 1984 were similar to those used during the 1983 studies,suitabil ity criteria specific to the lower reach can be developed.The purpose of this appendix is to verify the applicability of the suitability criteria developed in 1983 by Suchanek et al.(1984)for use in the lower river habitat studies.The general philosophy was to use the 1983 middle river criteria curves for the lower river unless the 1984 studies in the lower river provided evidence for modifications. METHODS The field sampling methods used are detailed in Section 2.1 of this report.These methods are very similar to those used during the 1983 studies (Suchanek et al.1984)and will only be summarized briefly here. Sampling sites included:(1)20 habitat model sites which were normally sampled twice a month and (2)31 opportunistic sites which were usually sampled only once. At each site,6 ft x 50 ft rectangular cells were sampled for fish and then habitat variables were measured in each cell.Cells were selected randomly at model sites,although sometimes additional selected cells with "goo dll habitat were also sampled.At opportunistic sites,cells were selected to encompass a variety of habitat conditions within potentially usable habitat.Habitat measurements taken at each cell sampled included a representative depth,mean column velocity,and estimates of primary cover type and percent cover (Appendix Table A-I). The data collected were examined for suitability criteria development by using the procedures described in Suchanek et ale (1984),with a few modifications. A-I Suitability was represented by mean catch per cell for chinook and coho salmon and proportional presence (proportion of cells sampled in which ~I 8fi@I Appendix Table A-I.Percent cover and cover type categories. Group #%Cover Group #Cover Type 1 0-5%1 No object cover 2 6-25%2 Emergent vegetation 3 26-50%3 Aquatic vegetation 4 51-75%4 Debris or deadfall 5 76-96%5 Overhanging riparian vegetation 6 96-100%6 Undercut banks 7 Gravel (III to 311 di ameter) 8 Rubble (3 11 to 511 diameter) 9 Cobble (larger than 51l diameter)- fish were captured)was used as the suitability measure for chum and sockeye salmon.Data were pooled by species for analysis.Some data were excluded from analysis by using results from the distribution and abundance analysis (Section 3.2)which indicated factors other than the microhabitat variables of velocity,depth,and cover were greatly affecting distribution.Macrohabitat type and turbidity were two factors which greatly affected distribution and were used as a basis for excluding cells fished.Cells which were excluded from the analysis varied by species and are detailed in the results section.The beach seine and electrofishing data were pooled for analysis because these sampling methods were both thought to be equally as effective given the sampling conditions.Although sampling efficiency varies by gear type and conditions fished,we assumed equal efficiency under all conditions as analysis of sampling efficiency was beyond the scope of this study. Groupings of habitat variables were identical to those used in 1983. Percent object cover categori es 76-95%and 96-100%were pool ed because of small sample sizes.Velocity and depth were pooled in groups identi- cal to those used in 1983 with the exception that cells with depths of 0.1 feet were examined separately.In 1983,only two cells with a depth of 0.1 feet were sampled,and therefore insufficient data were available for examination of suitability of this depth. Comparisons of the 1983 data with the 1984 data were made by plotting the suitabi 1ity criteri a derived in 1983 on the same graph with com- parable 1984 data.On the depth and velocity graphs this was done by normalizing the suitability to 1.0 for the 1984 depth or velocity increment with the highest suitability and then plotting the 1983 suitability criteria normalized to the same scale.The 1984 percent cover data were first regressed against catch per cell or proportional presence,and,if significant,the regression line was plotted and the suitability normalized to 1.0 for the highest cover category.The 1984 percent cover suitability line was then plotted on the same graph,by using the normalized 1.0 as the starting point.The suitability of A-2 - ~l - ,..., - ".... cover type for each species was calculated with the 1984 data using the methods described in Suchanek et al.(1984).The suitabilities cal- culated were then graphed against the cover type suitabilities calcu- lated in 1983. Variations in histogram distributions are to be expected on.a ,univariate basis given that percent cover,cover type,velocity,and depth together affect suitabilities of a cell.Therefore,composite weighting factors were calculated for each cell using the 1983 suitability criteria and .revi~ed 1984 criteria and then these weighting factors were compared with catch.Composite weighting factors were calculated by multiplying suitability indices for cover type,percent cover,and velocity togeth- er.For chinook and coho salmon,Pearson correlation coefficients were calculated between composite weighting factors and catch per cell (transformed by natural log (X +1)].Chi-square association tests were run between chum and sockeye proportional presence and composite weight- ing factor value intervals calculated using the 1984 criteria data. Intervals of composite weighting factors were specified by dividing the data into four groups of approximately equal sizes by value of the composite weighting factor.Pearson correlation coeffi ci ents and results of the chi-square analysis were then compared with the same analyses done in 1983.Most of the statistical tests and data manipu- lations were done with the Statistical Package for the Social Sciences (SPSS)(Nie et al.1975). If the fit of the 1984 data to the 1983 suitability criteria did not seem close upon visual inspection,the 1983 criteria were modified.One of the procedures for modifi cation was as foll ows.If,for exampl e,the 1984 velocity distribution data appeared to match closely the 1983 velocity criteria,the 1983 velocity criteria were input as suitabil- ities and averaged over each increment ofa variable such as depth for which a modification of suitability was desired.These averages were then multiplied by the mean catch of fish per cell divided by the mean suitability.The actual mean catches per cell by depth increment were then divided by the adjusted mean velocity suitability.If this ratio was less than 1.0,this would indicate less use of a depth increment than expected,given the average suitability for velocity.If the ratio was greater than 1.0,the use would be more than expected by adjusting for the effect of velocity.Sometimes this procedure..would be effective in taking out variation caused by the other variable.If necessary, this procedure was used to adjust for effects of two or more variables. If the above procedure was not effective in discounting the extraneous variation,then the criteria were modified using professional judgement. Correlations or chi-square association tests were then calculated between mean catch and calculated composite weighting factors using the modified criteria. RESULTS Abundance and distribution data (Section 3.2)have shown that the number juvenile chinook,coho,chum and sockeye salmon was very small at side sloughs in the lower reach.Even sampling cells at sloughs with good A-3 habitat failed to have any significant number of fish present in compar- ison with similar cells at the other macrohabitat types (tributary mouths and side channels).Fish were therefore responding to factors other than the availability of suitable microhabitat in their use of sloughs.For this reason,data collected at sloughs were eliminated from suitability criteria analyses to avoid comparing similar cells with large differences in mean catch. Chinook Salmon Chinook salmon suitability criteria were developed for both clear «30 NTU)and turbid (>30 NTU)water in 1983 because the catch in cells without object cover was much greater in turbid water than in clear water (Suchanek et ale 1984).Data collected in the lower river in 1984 have shown that turbidity may limit the distribution of chinook salmon by being too high (Figure 10).Since cells with good habitat were sampled when high turbidity was limiting use by chinook salmon fry,we decided to eliminate sampled cells with turbidities greater than 350 NTU. After-eliminating cells in side sloughs and cells with turbidities greater than 350 NTU,1155 cells were available for analysis of chinook distribution.Of the 1155 cells,400 were sampled in water with a turbidity of 30 NTU or less.Mean adjusted catch (catch adjusted to a standard cell size of 300 ft 2 )per cell of chinook fry in the 400 clear water cells was 1.3,while mean adjusted catch per cell in the 755 turbid cells was 1.1. A scatter plot of chinook salmon catch in cells without object cover versus turbidities ranging to 100 NTU was examined.No notable inflec- tions in catches of chinook salmon fry were noted over this range, although gradual increases in catches occu rred across the range.It seemed reasonable,therefore,to keep the same 30 NTU breakpoint between high and low turbidity data for this year1s analysis. Clear Water Correlations among the values of habitat attributes and clear water «30 NTU)chinook catch range to 0.32 in absolute value and a number of the correlations are statistically significant (Appendix Table A-2).In addition to these data,partial habitat data were recorded for four additional clear water cells and these additional data are used in subsequent analyses. Composite weighting factors for all cells sampled were calculated by using the 1983 suitabil ity criteria and al so with modification of the velocity criteria as proposed by EWT&A (1985)and then correlated with chinook catch transformed by natural log (x +1).In clear water,the correlation in 1983 was 0.43 but the correlation with the 1984 data was only 0.31 for the original criteria data and 0.26 with the change in velocity criteria proposed by EWT&A (1985).It was therefore deemed desirable to modify t~e criteria where large differences in individual criteria were found. A-4 - -- ~, - - ~ Appendix Table A-2.Kendall correlation coefficients between habitat variables and chinook catch by cell (N=396)for all gear ,~types,in clear water. f"""" Percent Cover Cover Type Velocity Depth Chinook Percent Cover 1.00 Cover Type 0.08*1.00 F'"'Velocity -0.32**0.04 1.00 .Depth 0.03 -0.08*-0.04 1.00 Chinook 0.07 0.09*-0.09*0.21**1.00 ....*Significantly different from 0 at p <0.05 . **Significantly different from 0 at p <0.01. -- -A-5 Least squares regressions were run between chinook catch per cell and the percent cover categori es in cl ear water.There was a si gnifi cant positive regression which is very similar to the suitability line developed in 1983 when the Y axis is normalized to a suitability of one (Appendix Figure A-I).The 1983 suitability criteria was therefore retained as a good estimate of this relationship. The distribution of mean catch per cell of chinook fry by velocity interval in clear water in 1984 shows that peak catches were made in sampling cells with a velocity ranging from 0.1 to 0.3 fps (Appendix Figure A-2).After normalizing this peak in catch to a sui·tability of- 1.0 and then plotting the 1983 suitability criteria on the same graph, it appears that chi nook used lower velocity water in the lower reach than in the middle reach under clear conditions.It was noted that the 1984 clear water distribution of catch by velocity interval was more similar to the 1983 turbid water velocity suitability criteria and therefore the 1983 turbid velocity criteria were plotted against the 1984 data (Appendix Figure A-3).Since the two distributions were similar,the 1983 turbid water velocity criteria were taken as a good estimate of the lower river velocity suitability for chinooks in clear water. Cover type suitabilities derived in 1984 for juvenile chinook in clear water contrasted sharply with those derived in the middle reach in 1983 (Appendix Figure A-4).Debris was used less by chinook in the lower reach for cover and emergent vegetation was used more.The sample size of the cobble/boulder cover category was only one and therefore this cover type could not be evaluated.Catches in the cells without object cover were also relatively higher in 1984 than in 1983. Therefore,it appeared that 1983 suitabi 1 ity for cover types woul d not apply in the lower reach.By adjusting for the effects of velocity and percent cover,better estimates of cover type suitability for the lower river were formulated from the 1984 data (Appendix Figure A-5).Since cobble and boulder sample sizes were low,suitabilities for these cover types were kept proportional in suitability to large gravel as was the case in 1984.Since the "no cover"catches were relatively large because fish were using relatively deep cells without object cover (see next paragraph),we lowered the suitability for no cover cells to 0.10, the suitability found in 1983. A heavy use of deep,clear water by chinooks was found in 1984 while in 1983 the data suggested a peak in use of cells 1.0 to 1.5 feet deep (Appendix Figure A-6).In 1983,an evaluation of depth found it had little effect on increasing the correlation of fish catch with composite weighting factors using it.Depth was used in the 1983 modelling efforts as having no value if less than 0.14 ft and having a suitability of 1.0 if greater than 0.15 ft.In order to evaluate depth,sUitability criteria were fit to the data using professional judgement after first adjusting for mean velocity and percent cover suitabil ity (Appendix Figure A-7). After the modifications to the cover suitability and depth criteria were made,we then correlated transformed chinook catch with the composite A-6 - - - -- - 18.9 10.0 I .:t STANDARD ERROR N =NUMBER OF CELLS SAMPLED -=1983 --=1984 ~ 0 0 Z 8.0-~ 0 ..f ..J I.LI 6.00..... :r: 0 ~ c( 0 Z 4.0 c( I.LI:e 2.0 1.0 X I.LI0.8 0z 0.6 )- ~-..J 0.4 CD c( ~ 0.2 :; (I) 0.0 --'-...L-.__....L-....L-__.l.--.l.--_---I"---I_~-..L-..L-~--'--"-0.0 N-168 N=125 N-61 N-34 N-12 (0-5%)(6-25%)(26-50%)(51-75".)(76-100%) PERCENT COVER CATEGORIES Appendix Figure A-I.Mean catch of juvenile chinook salmon per cell by percent cover category (bars)in clear water of the lower Susitna River,1984 and comparison of fitted suitability indices (lines)calculated in 1984 and for the middle Susitna River,1983. A-7 - e.2 Ha8S :0::I!.STA~DARD ERROR0S.o0 Z N -NUMf!ER OF CELLS SAMPLED J:---1983 1.0 u 4.0.X ..J 0.8 ~~..J Z1&1 3.0 U 0.6 >-.... J:I- U 2.0 0.4 ~I-et CD U et Z 1.0 I- <0.25 1&1 (/) ~H:a;'4 0.0 0.0 0.0 0.3 0.6 0.9 1.2 1.!5 1.8 2.1 3.4 VELOCITY (U/••cl Appendix Figure A-2.Mean catch of juvenile chinook salmon per cell by velocity intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River, 1983. 7.0 I~STANDARD ERROR :0::N-88 0 Ii.O N-NUMBER OF CELLS SAMPLED0 Z -aI9B3(TURBIO) J:!l.0U 1.0 ..J X..J 4.0 1&1 1&1 u 0.8 0....z J:3.0 U 0.1i >-I-<I- U 2.0 ..J Z 0.4 ai et < lI.I 1.0 I- ~0.2 ::; (/) N s t4 0.0 0.0 0.0 0.3 0.6 0.9 1.2 1.!5 1.8 2.1 3.4 VELOCITY (ft/sec I Appendix Figure A-3.Mean catch of juvenile chinook salmon per cell by velocity intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for turbid water in the middle Susitna River,1983. A-8 - ,.- o _CHINOOK,1983 VZ2I CHINOOK,I984 N =NUMBER OF CELLS SAMPLED <D II Z x ~cz 0.25 1.00 0.75 >- ~-0.50 ..J OJ <t ~ ::> (/) ,-0.00 tJ)~¥'0:1.&1 OZ I.&I..J (!)Z ~Z a: a:jZ 1.&11.&1 ..J -0 (!)w zet ZO 1.&1 oet ..Jo m ~-o:~tLI->CD a:m m.J m et~(!)a:(!)!;i 01.&1 1.&1 03::>::>::>~eta:Zet a:~00.J(!)etQ.Q 00 0:01.&1 %-1.&11.&1 0Zomet(!)0:0:2!C>z::>1.&1 1.&1 ILl ILl>>> 0 COVER TYPE Appendix Figure A-4.Comparison of cover type suitability indices for juvenile chinook salmon in clear water calculated from 1984 lower Susitna River distribution data and 1983 distribution data. A-9 -....--N,.:..!.!...O ..- - - - - -N=87 r-- -N=14~ N=I....--N=18-..- N=71 r--- -N=62~ 1.00 0.80 X LIJ Q 0.60z ,....-..J-OJ 0.40 oct...-~ U) 0.20 0.00 N=6 N=NUMBER OF.CELLS SAMPLED N=31 - - en 1-:lI:::......0:11.1 UZ 1&.I..J ~z I-Z 0: 0::::)z lIJlI.I ..J -0 ~1&.1 z«zo 1&.1 In 0«..Jc m !i-0:>~o::1&.1-> 1&.1 O:lD ID..J In :::)!i et«z«(!)I-0 C 1&.1 m:J :::)01-..JO:0:«0c000:c!>«a.11.11-z «11.1 x-:::lEu.0:::)om ~0:0:1&.1~Z11.1 11.1>>> 0 COVER TYPE Appendix Figure A-5.Cover type suitability indices for juvenile chinook salmon in clear water calculated from 1984 lower Susitna River distribution data after adjusting for velocity and percent cover. A-IO --)j )-))1 --]<I 1 1.0 0.6 >- I- ..J 0.4 iii ~ 0.2 :; C/) X LLI 0.8 az 4.5 N=36 N=55 I±STANDARD ERROR N·NUM8ER OF CELLS SAMPLED -=1983 O 0 i I -,I !!I\,I I 00. I I I I I I I I I I II . 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 DEPTH (ft) 1.0 12.7 9.0 ~o 8.0oz:E 7.0 o •..J 6.0 ..J LLIo 5.0 ...... :ro 4.0 ~o 3.0 Z<tLLI 2.0 ~ ;p I.............. Appendix Figure A-G.Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in clear water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983. N=36 0.6 >- t- ...J 0.4 CD <tt- 0.2 :5 CJ) 1.0 X iJJ 0.8 0z - - -....l\.r N=!55 /' /' I / I I I Ir--------f--F-J !0=,30 I N:127 N:144 I I±STANDARD ERROR N:NUMBER OF CELLS SAMPLED --:1984 12.7 9.0 ~ 0o 8.0 z-::I:7.0u I ...J 6.0 ...J W U 5.0 ....... ::I: U 4.0 )::a I ~ I 1-'"U 3.0 N Z <t W 2.0 ~ 1.0 4.5 00"""-I I I I ~OO. I i I I I I I I «I I . 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 DEPTH (ft) Appendix Figure A-7.Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in clear water of the lower Susitna River,1984. Suitability index (line)fitted by hand usin9 professional judgement. 1 J j i J )J ~J ]I j J I J 3 I weighting factors calculated with the 1983 percent cover criteria and turbid water velocity criteria along with the 1984 lower river cover type and depth suitability criteria.The correlation was 0.61, substantially higher than the original 1983 criteria.If depth was eliminated from the calculations,the correlation dropped to 0.26 and if primary cover type was dropped the correlation dropped to 0.52.There- fore,it seemed reasonable to keep the newly modified cover type and depth criteria as inputs. Turbi d Water Correlations between the values of habitat attributes and chinook catch in turbid water range to 0.39 in absolute value and a number are statis- tically significant (Appendix Table A-3).Partial habitat data were recorded for 11 additional turbid cells and these additional data were used in subsequent univariate histograms. Correlations between composite weighting factors calculated with the 1983 turbid water criteria and 1984 chinook catch was 0.31,wh"ile composite weighting factors calculated by incorporating the cover modifications proposed by EWT&A (1985)were correlated with an r-va1ue of 0.26.Comparable correlation with the 1983 data was 0.38.These data again suggested that some modifications could be made,especially given the changes already made °in the clear-water cover type suitabil- ities. A comparison of 1984 velocity distribution data and the 1983 velocity suitability criteria for chinook salmon showed few differences (Appendix Figure A-8),and therefore the 1983 velocity criteria were accepted as the 1984 criteria curve. Least squares regressions were run between chinook catch per cell and the percent cover categories in turbid water.There was no significant relationship between catch per cell and percent cover category and mean catch per cell decreased with increases in cover (Appendix Figure A-9). By adjusti ng for vel oci ty,a s1i ght trend upwa rd was noted over the first three categories.The percent cover criteria developed in 1983 was therefore accepted as reasonable,as increases in the amount of object cover would seem more desirable for fish and sample sizes were very small in the 51-75%and 76-100%cover categories. In 1983,cover type for chinook in turbid water was not evaluated. EWT&A (1985)modified the turbid water criteria,however,so that they more closely reflected the clear water criteria developed in 1983.In 1984,mean catches of chinooks in turbid water were highest in the emergent vegetation,rubble,and debris-deadfall categories,but catc~es were only slightly higher than in the cover category "no cover". Cover type was evaluated in 1984 by using the method of EWT&A (1985)for calculating turbidity factors from the fitted regressions of percent cover in clear and turbid water and their associated chinook mean catches.Turbidity factors were calculated (Appendix Table A-4)and then applied to the revised lower river cover suitability data.These A-13 ""'", ~ Appendix Table A-3.Kendall correlation coefficients between habitat variables and chinook catch by cell (N=744)for all gear types~in turbi d wa ter.~ - Percent Cover Cover Type Velocity Depth Chinook ~ Percent Cover 1.00 Cover Type 0.39**1.00 Velocity 0.05*0.16**1.00 I!'!A': Depth 0.06*0.26**0.21**1.00 ..".,. Chinook -0.02 0.00 -0.17**-0.15**1.00 *Significantly different from 0 at p <0.05. **Significantly different from 0 at p <0.01. ""'"I, - - A-14 f"""2.5 N-20T I:!:STANDARD ERROR N·NUMBER OF CELLS SAMPLED ~,---1983 1.0 ¥2.0 0 0 Z ~ Q O.B 1.5 )(.J .J IU IU 0 Q Z,...,....0.6 :r 11-39 >- Q 1.0 N'1I2 ~~ c(.J Q 1II Z NoS2 0.4 c( r~c(~ IU 5 :IE 1II 0.5 0.2 11-97 0.0 0.0 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 4.8 VELOCITY (ft/sec) .- Appendix Figure A-8.Mean catch of juvenile chinook salmon per cell by velocity intervals (bars)in turbid waters of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983. 2.0 lIl: 0 0 Z U5::c 0,-.., ..J ..J W 1.00 "-::c 0... ~ 0 0.5 Z ~ lIJ ::I 0.0 ;F I~STANOARD ERROR N-NUMBER OF CELLS SAMPLED ---1983 N-525 N'161 N-46 (0-5%)(6-25-1.){26-50%1 (51-7S%1 (76-10<W.) PERCENT COVER CATEGORIES 1.0 Xw0.8 0z 0.6 ~ !:: 0.4 ...J ai ~ 0.2 !:: ::I U) 0.0 Appendix Figure A-9.Mean catch of juvenile chinook salmon per cell by percent cover category (bars)in turbid water of the lower Susitna River,1984 and fitted suitability index (line)calculated for the middle Susitna River,1983. A-IS revised suitabilities were much too low for many categories given observed catches and therefore a suitability of 0.15 was assigned as a minimum for cover type suitability in turbid water based on observed mean catches.Using this method,none of the suitabil ities for cover type in conjunction with percent cover in turbid water are greater than 0.40 (Appendix Figure A-10). Appendix Table A-4.Calculations of turbidity factors for 1984 lower river data. Number of Fish Per Cell (Fitted to a Line Percent Turbidity Cover Clear Turbid Factor 0-5% 6-25% 25-5m~ 51-75% 76-100% 0.5 1.5 2.5 3.5 4.5 1.1 1.3 1.5 1.7 1.9 2.2 0.9 0.6 0.5 0.4 ~I In turbid water,peaks in chinook use were found in water less than 0.5 ft deep in both 1983 and 1984 (Appendix Figure A-ll).In 1983,since fitting the depth suitability line to the data did not increase the composite weighting factor much,the depth criteria used for clear water (0 if less than 0.14 ft,1.0 if greater than 0.15 ft)was used for modelling. In 1983 there was only one turbid cell sampled with a depth of 0.1 feet and therefore the value of cells with this depth could not be evaluated. For purposes of IFIM modelling,this depth was assigned a suitability of 0,while in the RJHAB model data this depth did not occur.In turbid water,21 cells of 0.1 feet depth were fished in 1984 and the mean catch was 0.5 chinook juveniles per cell.These data suggest that under turbid conditions the value of 0.1 feet cells is greater than o.A suitability criteria line was fit to the 1984 turbid water depth data by first adjusting for the effects of velocity (Appendix Figure A-12).The optimum depth"ranged from 0.3 to 1.5 feet. Once all the criteria were modified,correlations were calculated between catch transformed by natural log (x +1)and the composite weighting factor calculated by multiplying the suitabilities for ve- locity,cover,and depth together.The correlation was 0.33,and if depth were removed the correl ation dropped to 0.28.If cover was removed from calculations of the composite weighting factor,the corre- lation increased to 0.36.Since instream cover has value as a velocity break in turbid water,it seemed reasonable to keep velocity,cover,and depth in the modelling. A-16 ~\ - .... --)( llJ 0.4-0 Z >-- ~ ..J lD 0.2- ~ ~ ::I -en 1""". PERCENT COVER .1 0 -~ .2 6 -25 .3 26 -~O .4 ~I -7~ .~76-100 PERCENT COVER BY COVER TYPE Appendix Figure A-lO.Cover type suitability indices for juvenile chinook salmon in turbid water developed from 1984 lower Susitna River chinook turbid water distribution data. 1.75 ¥o 1.50o !:z:u 1.25 ..J ~1.00 U.... :z:0.75 u t- <{ u 0.50 z <{ ~0.25 T I!STANOARD ERROR N-NUMBER OF CELLS SAMPLED ---1983 1.0 )( 0.8 ~ Z 0.6 >- l- ..J 0.4 iii <{ t- 0.2 :; en 4.3~: 0.00 +...J--r--r-~-"",-""".....,r--r-""'-,.-.....,rJ...-r-..-J\r-+..J..0.0 0.0 0.2 0.4 0,6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 DEPTH (ftl Appendix Figure A-II.Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in turbid water of the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983. A-17 \.0 Xw0.8 0 z- 0.6 >- ~ ....I 0.4 m oct ~ 0.2 :5 en 0.0 LED -N=1I6 It STANDARD ERROR N=252 - -N=220 N-NUMBER OF CELLS SAMP --"1984 ,- ,-r--1-----_.1----~ \r-I 1\I-I l.- e-I \-I \,-I \r·I \N=61 ,... I \...I- II -----Jy-....r- j N=76 r-- I T t\I J,.t..'If 1 .... V \IIII.0.00 I I I I I I I 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.3 DEPTH (ft) 1.75 ....I....I 1.00 UJu .....075X. <.) ~ctU 0.50 z ct LLJ 0.25 ~ ~1.50 o Z :i:1.25 u :too I........ ex> Appendix Figure A-12.Mean catch of juvenile chinook salmon per cell by depth intervals (bars)in turbid water of the lower Susitna River,1984. Suitability index (line)fitted by hand using professional judgement. »._I J J , ••J I .t J 1 il J ,I -~ - - - - - - Coho Salmon Juvenile coho salmon suitability criteria were developed only for clear water in 1983.Very few coho were captured in macrohabitat types other than tributary mouths in the lower reach and therefore only tributary mouth data were used in suitab-llity criteria comparisons.Most of the turbidities in the'tributary mouths were less than 30 NTU although on two occasions,turbidities were over 100 NTU. A total of 345 cells with complete habitat data were sampled in tribu- tary mouths and another 2 cells with partial habitat data were sampled. Mean adjusted catch in the cells sampled was 1.2 fpc.Kendall corre- lations among the values of habitat attributes and coho catch ranged to 0.43 in absolute value (Appendix Table A":5).Cover type was most highly correlated with coho catch. The distribution of mean coho catch per cell by velocity interval in 1984 matched quite closely with the suitability criteria derived in 1983 for the middle river (Appendix Figure A-13).The 1983 velocity criteria were therefore chosen as representative for the lower river. A regression of coho catch to percent cover category was significant (Appendix Figure A-14).When the 1983 and 1984 data were normalized to 1.0 on the V-axis for the 76-100%category,the 1983 SUitability line had a much greater slope,and suitability for 0-5 percent cover in 1983 was 0.12,while in 1984 it was 0.33.After adjusting for the effect of velocity,the distribution of catches by percent cover interval appeared to be more similar to the 1983 distribution and since the sample size in 1983 was larger,the 1983 percent cover suitability relationship was chosen for use in the lower river. Initial calculations of the suitability of cover type for coho salmon indicated that suitabilities in the lower river were similar to those found in 1983 (Appendix Figure A-15).After adjusting for the effects of velocity and percent cover,these estimates of cover suitability for the cover types were revi sed for use in the lower ri ver in 1984 (Appendix Figure A-16).Since sample sizes for the three substrate cover types were small,the suitability of 0.10 calculated in 1983 for rubble and boulders was used for these three categories. The distribution of CPUE 's for depth was very different from that found in 1983 (Appendix Figure A-I?).By adjusting for the effects of ve- locity,percent cover,and cover type there still was no trend in depth suitabilities and therefore depth suitability was not changed from that used in 1983. The correlation between transformed coho catch and the composite weight- ing factor calculated by multiplying the velocity,cover,and depth suitabilities together was 0.32. Sockeye Salmon Juvenile sockeye salmon suitability criteria were developed by pooling data over gear type and turbidity level in 1983.Since abundance and distribution data have indicated that sockeye salmon use of lower river A-19 Appendix Table A-5.Kendall·correlation coefficients between habitat variables and coho catch by cell (N=345)in clear water. A-20 - - - - N~35 T~STANDARD3.0 ERROR J.. 0 N-NUMBER OF CELLS SAMPLED::z::2.5 N=&7 -.-)9B30 0 1.0 ...J 2.0 X...J kJ O.B kJ 0 Q "-1.5 N-9 :!!:-%"'0147 0,60 ~ ~!:c:(1.0 ...J00.4 ai Z c:( ~0.5 0.2 !: ::Iii:;:) en 0.0 0.0 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.2..-VELOCITY (tt/,.c) Appendix Figure A-13.Mean catch of juvenile coho salmon per ce 11 by vel oei ty i nterva 1s (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983. 3.5 o 3.0 Xo o 2.5 I!STANDARD ERROR No NUMBER OF CELLS SAMPLEO --01983 --01984 1.0 )( 0.8 kJ Q Z 0.6 ~ ~ 0.4 ...J iii 0.2 ~ ;:) en 0.0 1 ...J ~2.0 o "- %1.5o ~ C.:J 1.0 Z c:( ~0.5 0.0 -l..~N~_';';13;-:;7""'''''''7.N~0'::'8;-1L..J-N::"6::':6;:-l-~N::-0--4:-:0::-L...l..:N::-o-:2~2...L....l- (O·S'IIo1 (6-25'110)126-50'1101151-75'1101 (76-100'll01 PERCENT COVER CATEGORIES - Appendix Figure A-14.Mean catch of juvenile coho salmon per ce 11 by percent cover category (bars) in the lower Susitna River,1984 and comparison of fitted suitability indices (lines)calculated in 1984 and for the middle Susitna River,1983. A-21 ~, -, -<D II 1.00 z iWll COHO,1983 ~COHO,1984 N =NUMBER OF CELLS SAMPLED ~, 0.75 - X LLI C Z >-....0.50 ~ -I ID-alcs:....-::>en 0.25 - 0 ID "Z ~ 0.00 en ~~.....0:::ILl oz ILI-J C)Z ~z a: a:::::>z ILl ILl -J -0 C)ILI z«zO ILl0«-J c m ~-a:::>--ILl->m m-J «~C)a:::C)!;tl.LI a:::m m::::>m ::::>c:(««zc:(0 ~ILl ::::>at--Ja:::a:t-0c0000:::«ILl C)«4-WILlomx-0zC)a:::a::::lEC)::::>ILl ILl ILl ILl Z >>> 0 ~; COVER TYPE Appendix Figure A-IS.Comparison of cover type suitability indices for juvenile coho salmon calculated from 1984 lower Susitna River distribution data. - - A-22 ,- 1.00 0.80 x ~LLI 0.60c Z ~>-... oJ CD 0.40 1""'"t! ;:) en- 0.20 0.00 lD II Z-- - - co II-Z.--- It)-II-Z- 0-10 V 01 II (\/Z 11 r----Z r--- - 0 co =II II II-Z Z Z.----.--- 0-In II Zn !!!~~.....a::LU OZ LU~<.!)Z ~z a:: a::~z WlIJ ~-0 C)lIJ z<zo LU0<~o ID ~-0::>LU->ID O::ID CD~ID <I-«(!)a::ClI-oW~<z<a::<0 lIJ CD;:)~a~~a::0 0 00 a::<a-wI-z Om <lIJ (!)z-:::l!!E LU 0 ~(!)a::a::LU~zLULU>>> 0 COVER TYPE Appendix Figure A-16.Cover type suitability indices for juvenile coho salmon developed for the lower Susitna River in 1984. A-23 ~205j I±STANDARD ERROR N"NUMBER OF CELLS SAMPLED o 2.0 N=131 -=1983 I ..J N=26..J X ~1.5 1.0 ~ ......0.8 z:I:o 1.0 >-I-0.6 I-«-0 ..J0.4 -)::-I z 0.5 CDI«N «+::0 lLJ 0.2 I- ~0.0 :::>0.0 U) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.5 DEPTH (ft) Appendix Figure A-17.Mean catch of juvenile coho salmon per cell by depth intervals (bars)in clear water of the lower Susitna River, 1984 and fitted suitability index (line)developed for the middle Susitna River,1983. J ,~J J J J l t J I ..~i J J ]J side channel s is 1 imited by hi gh turbidities (Figure 18),cell s with turbidities greater than 250 NTU were e1 iminated from suitabi1 ity criteria development. After cells with turbidities greater than 250 NTU were eliminated,922 cells with complete habitat data were available for analysis.Sockeye were captured in 117 (12.7%)of these cells.Correlations among the habitat variables ranged to 0.65 in absolute value and velocity was most highly correlated with sockeye catch (Appendix Table A-6).In addition to these cells,partial habitat data were collected at six additional cells and these data are used in subsequent univariate histograms. The distribution of proportional presence by velocity interval was very similar to that found in 1983 (Appendix Figure A-18).There was no use of velocities greater than 1.2 fps,however,and in 1983 there also was no use of velocities greater than 1.2 fps although sample sizes were smaller.Since these high velocities are not used,the lower river velocity suitabi 1 ity criteria were modifi ed so that velocities greater than 1.2 fps have 0 suitability (Appendix Figure A-18). Distribution of proportional presence by percent cover categories was similar to that found in 1983 (Appendix Figure A-19).The.1983 suita- bility relationship was therefore selected for use in 1984. The distribution of proportional presence by cover type categories was somewhat different than that found in 1983 (Appendix Figure A-20). Suitabi1 ities for the cover types used in the lower river in 1984 w"ill be those developed in 1984 wi th the fo 11 owi ng two exceptions.Si nce sample sizes were small (less than 25)for the cover type categories, undercut banks and overhanging riparian vegetation,the suitabilities calculated in 1983 were averaged with the 1984 suitabilities to give a value intermediate between the two. No trend was noted in the 1984 depth distribution data and therefore no suitability criteria were fit to these data (AppendiX Figure A-21).Of the 20 cells sampled with 0.1 ft depth,fish were sampled in 2 suggest- i ng that thi s depth does have value.Therefore any depth wi 11 be assumed to have a suitability of 1. Composite weighting factor intervals calculated by multiplying cover and velocity suitabilities together were associated with proportional presence of sockeye salmon (AppendiX Table A-7). Chum Salmon Juvenile chum salmon suitability criteria were deve1qped by pooling data over gear type and turbidity in 1983.Abundance and distribution data indicate that chum salmon use of lower river side channels is limited by high turbidities (Figure 15).Cells with turbidities greater than 200 NTU were eliminated from sUitability criteria development.Also,since most chum salmon outmigrate before July 16,only data collected before this date were retained for sUitability criteria analysis. A-25 Appendix Table A-6.Kendall correlation coefficients between variables and sockeye catch by cell {N=922}. habitat Percent Cover Cover Type Velocity Depth Percent Cover 1.00 Cover Type 0.30**1.00 Velocity -0.18**0.65**1.00 Depth 0.05*-0.01 0.07**1.00 Sockeye 0.04 -0.06*-0.21**0.02 *Significantly different from 0 at p<0.05. **Significantly different from 0 at p<O.01.- 0.30 .,I:.STANDARD ERROR(ij M -,...z ~~0.25 N •NUMBER OF CELLS SAMPLED ...I<n -"1983 \,0lULU u~0.20 --·1984 (REVISED)X LLl -lL.0.8 0OLU~>-~LUO.15 I'l >-_lll::2 0.6 ......U 0::0 ...I ~<n 0.10 en 0:1:0.4 <l 0::)-)- Q.i 0.015 ~0.2 (I)""'"' 0.00 0.0 0.0 0.3 0.6 0.9 1.2 \,5 1.8 2.1 4.8 VELOCITY (ftlsee) -Appendix Figure A-lB.Proportion of cells with juvenile sockeye salmon present by velocity intervals (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983 and revised in 1984 for the lower Susitna River using professional judgement. A-26 ]"J ]j ))))'J 1 1 1 ) 1,00 ...z C1 SOCKE'I'E,1983 lZ2'SOCKEYE,1984 N •NUMBER OF CELLS SAMPLED ....z '''1 I :,"-::' 0.31:1.,I:t STANDARD ERROR '.~~\; )("t Oi...~:~.- 0 ~.z 0.30 N-NUMBER OF CELLS SAMPLED ~....z (l)Z -'198~::0.150...JIIJ ---1984iii(I)0.25 ::::i 01lJ iii0::1.0 ~J!l",,11.o 0.20 XIIJ (I').! Z>-0.8 ~~! I 01lJ 0.215 ;~)::-j::lIC 0.11:1 z I - N 0::0 0.6 >- -....J 0 0 11.(1)...o 0.10 :::io::Z 0.4 iiiQ.!:<l:..lli~0.01:1 ...0.00 .~ 0.2 :5 II)1-'""'0:....OZ ........1 t!>Z 1-2 ca:: iii:=>Z ............I ~2 <lI ....zo(ZO '"00(....10 III o::~;i2 ",->(I)III <lit(...."'Ill 111....1 III =>le 0(0::Z4 0 0 UJ Ill=>=>01-....It!>4Q.0::1-0 0.00 I I I I I I I ,I I I ,0.0 0 00 0::UJUJ4",:z:iii:~N'1:I31 N-233 N-IOI N'41 N·22 z 0111 <!I 0::Et!> (0-1:1%)(6-25%)(26-1:10%)(51-75%)(76-100%) =>w III wUJ >>> 0 PERCENT COVER CATEGORIES COVER TYPE Appendix Figure A-19.Proportion of cells with juvenile sockeye salmon present by percent cover category (bars)in the lower Susitna River,1984 and comparison of fitted suitability indices (lines)calculated in 1984 and for the middle Susitna River,1983. Appendix Figure A-20.Comparison of cover type suitability indices for juvenile sockeye salmon calculated from 1984 lower Susitna River distribution data and 1983 middle Susitna River distribution data. 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.5 DEPTH (ft) I±STANDARD ERROR N '"NUMBER OF CELLS SAMPLED N=131 -=1983 ):a I N 00 r-t/)Z 0.20 ...I~ ...It/) ~~lOXoIfO.I 5 N=307 N=302 N:90 r+'.~ ~ZOLLJ0.8_ ~~O.IO 0.6~ r-~_ 0:g 0.4 I 0(/)0.05 m ~:J:0.2 ~Q:r-5 a..~0.00 0.0 t/) 0.0 Appendix Figure A-21.Proportion of cells with juvenile sockeye salmon present by depth intervals (bars)in the lower Susitna River,1984 and suitability index (line)developed for the middle Susitna River,1983. I )J j t j J )1 I )..~,)]})J -Appendix Table A-7.Proportional presence of sockeye salmon associated with the composite weighting factor calculated by multiplying velocity and cover suitabilities to- gether. Composite Weighting Total Number Proporti on With Factor Interval of Cell s Fish Present Chi-Square o -0.06 244 0.02 X2 =55.3.....0.07 -0.11 213 0.08 0.12 -0.19 228 0.17 p<O.OOl 0.20 -1.00 241 0.23 :1f1Iii'i'II!A, The number of cells available for analysis of juvenile chum distribution totaled 249 after elimination of the cells outlined above.Chum salmon were captured in 98 (39.4%)of these cells.Correlations among the habitat variables and chum fry catch ranged to 0.32 in absolute value (Appendix Table A-8).Partial habitat data were collected at two additional cells. .- ..... - - The chum salmon distribution by velocity interval in 1984 was similar to 1983 (Appendix Figure A-22).Therefore,the suitability criteria for chum salmon developed in 1983 was selected for use in 1984. In 1983,the relationship of chum salmon use to percent cover and cover type was the weakest of any of the four species.In 1984,the 0-5% cover category and the "no cover"type had the highest proportional presence within their respective distributions (Appendix Figures A-23 and A-24).These data indicate that chum salmon fry do not orient to cover during rearing.Even when velocity suitability was adjusted for, no real trends in percent cover and cover type utilization were noted, although large gravel and rubble were used sl ightly more than was the "no cover"type.Since there were no trends,cover type and percent cover will not be used in the 1984 analysis of chum habitat use. The distribution of chum proportional presence by depth intervals in 1984 was similar to that found in the 1983 studies (Appendix Figure A-25).Since the distributions were similar,the criteria fit in 1983 was used to test for the value of depth in increasing the associations with chum catch.Therefore velocity was first used alone and then with depth to form categories which were associated with chum proportional presence. Although composite weighting factors calculated by velocity alone and velocity and depth together were both significantly associated with chum proportional presence,the composite weighting factor calculated by depth and velocity together seemed to fit the observed distribution data better (Appendix Table A-9).Therefore both velocity and depth suita- bility criteria will be used to model chum salmon habitat. A-29 Appendix Table A-B.Kendall correlation coefficients between habi tat variables and chum catch by cell (NzZ49)for all gear types,turbidity below 200 NTU.- ..... Percent Cover Cover Type Velocity Depth Chum Percent Cover·1.00 Cover Type 0.13·...1.00 Velocity -0.25.....0.15**1.00 Depth -0.05 -0.03 0.07 1.00 ...., Chum -0.20**-0.07 -0.04 -0.32*·1.00 ..Si gnificantly different fl"Olll 0 at p<0.05•-Significantly different from 0 at p<O.OL -- ~, - I~STANDARD ERROR N '"NUMBER OF CELLS SAMPLED -·1983 N-or IIz 1.0 1.0 X ILl 0.8 ~ 0.6 >- l- 0.4 ;J a::l 0.2 ~ :5 0.0 -I.+,~-r--,.....-.....,...r--"""T..L-""'="""I'"'L-_....,..~_~.T\.r-+-....L.0.0 C/) 0.0 0.3 0.6 0.9 1.2 I.e 1.8 2.1 4.8 VELOCITY (ftlsec) Appendix Figure A-Z2.Proportien of cells with juvenile chum salmon present by velocity intervals (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River,1983.- A-30 - 0.0 -I±.STANDARD ERROR- (I)....0.5 -N =NUMBER OF CELLS SAMPLED r-\.0:lz -LlJLlJ I- ,.-0(1)0.4 -XLlJI-o.a LlJu..o::-~0OCL.-I-Z z:t 0.3 -I-0.62::::>>-....%... I-.... 0::0 - 0%0.2 --0.4 ..J-~....CD-I-et0::-....CL.~0.1 -I-0.2 ::::> f"""-en I- 0.0 0.0 N-119 N"'74 N-37 N-12 N - 9 j.t~(0 -5%)(0-25%)(26-50%)(51-75%)(76-100%) PERCENT COVER CATEGORIES A~pendix Figure A-23.Proportion of cells with juvenile chum salmon present by percent cover category (bars)in the lower Susitna River,1984 and fitted suitability index (line)calculated for the middle Susitna River,1983. -A-31 0.8 0.7 N=4 I~STANDARD ERROR N =NUMBER OF CELLS SAMPLED - 0.3 0.5 ...zau (/)aua:0.6 a.. ::E ::;) %:o (/) -I 0.4 -Iauo lLo zo...a:o 0..0.2oa:: 0.. 0.1 0.0 (J) 0::m LLIc N=29 N=53 COVER TYPE N=44 0::' LLI>8 oz Appendix Figure A-24.Proportion of cells with juvenile chum salmon present by cover type (bars)in the lower Susitna River,1984. A-32 ~, ~, - ..- I±STANDARD ERROR N'NUMBER OF CELLS SAMPLED -.1983 1.0 1.0 x w 0.8 ~ 0.6 ~ 0.4 :::! !D ~ 0.2 t 0.0 +Lr--,---.l.,..---r-.,J-r--.,.---.Jl.,---,--,Jc:::;:::±~::::::j....LO.0 ~ 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.3 DEPTH (ft) Appendix Figure A-25.Proportion of cells with juvenile chum salmon present by depth intervals (bars)in the lower Susitna River,1984 and fitted suitability index (line)developed for the middle Susitna River, 1983. Appendix Table A-g.Proportional presence of chum salmon fry associated with several composite weighting factors. A-33 Summary A summary table of reV1S10ns of the middle river suitability criteria for use in the lower river reveals that about half the criteria were not changed or changed only slightly (Appendix Table A-I0).The velocity and percent cover relationships were often not changed while the depth and cover type criteria have often been modified •.Point specific values for all the sultability criteria developed for use in the lower river are presented in Appendix Table A-II. DISCUSSION Chinook Salmon The turbid water velocity criteria developed in 1983 were used for both clear and turbid chinook distributions in the lower river in 1984.The reason that there was no shift in velocity optima from cl ear to turbi d water may be due to several factors.In the middle river,substrate is ·much larger and therefore,juvenile chinooks may find higher velocities because suitable as there is always some substrate cover to hide under or behind.In the lower river,however,very little substrate cover is present and therefore chinook use lower velocity water much more. In the lower river,cover suitabilities were often somewhat different than in the middle river.Part of this difference may be due to the actual cover in cover type categories being of a different type.For instance,the aquatic vegetation in Caswell Creek,which harbored large numbers of chinook fry,was not present in any of the sampled streams in the middle river.Also the debris cover type in the lower river was often much more silted in than in the middle river and therefore less suitable.The primary cover type is associated with a variety of secondary cover types and it is likely that,on the average,secondary cover types associated with a primary cover type in the lower river are different than the secondary cover types most common in the middle river.If these secondary cover types are more suitable for fish,then they might raise the suitability of the primary cover type. Most notable in the analysis of chinook suitability criteria was the effect of depth upon the distribution of chinook salmon.In the lower river,chinook salmon found deep,water m~ch more suitable than in the middle river (Appendix Figure A-7).This 1S probably due to the tribu- taries in the lower river having a turbidity of approximately 10 to 20 NTU and therefore depth mi ght have a cover value in deeper waters.In the middle river,much of the data were collected in Portage Creek, Indian River,and other areas where the turbidity was usually less than 5 NTU and depth would not provide cover at depths which can be sampled. Sometimes juvenile salmon thought to be chinook fry could be seen feeding on the surface in tributary mouths such as Rolly Creek where depths were greater than 5.0 ft. In turbid water,on the other hand,depths greater than 1.5 ft were less suitable than shallower cells (Appendix Figure A-Il).This trend was A-34 - - - - - - - - """ - - Appendix Table A-10.Summary of reV1Slons of 1983 middle river juvenile salmon criteria for use in the lower Susitna River, 1984. A-35 Appendix Table A-II.Suitability indices for juvenile salmon for velocity,depth,and cover in the lower Susitna River,1984. VELOCITY Chinook Coho Sockeye Chum Velocity Suita-Velocity Suita-Velocity Suita-Velocity Suita- (ft/sec)bi 1ity (ft/sec)bi 1ity (ft/sec)bil ity (ft/sec)bil ity 0.00 0.42 0.00 0.29 0.00 1.00 0.00 0.86 0.05 1.00 0.05 1.00 0.05 1.00 0.05 1.00 0.35 1.00 0.35 1.00 0.20 0.71 0.35 1.00 0.50 0.80 0.50 0.88 0.50 0.48 0.50 0.87 0.80 0.38 0.80 0.55 0.80 0.35 0.80 0.70 1.10 0.25 1.10 0.32 1.10 0.14 1.10 0.56 ):0 1.40 0.15 1.40 0.12 1.30 0.00 1.40 0.37, w 1.70 0.07 1.70 0.04 1.70 0.150\ 2.00 0.02 2.00 0.01 2.00 0.03 2.30 0.01 2.10 0.00 2.10 0.00 2.60 0.00 DEPTH Chinook (turbid)Chinook (clear)Coho Sockeye Chum Depth Suita-Depth Suita-Depth Suita-Depth Suita-Depth Suita- (ft)bil ity (ft)bil ity (ft)bi 1ity (ft)bil ity (ft)bil ity 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.10 0.29 0.15 0.00 0.14 0.00 0.10 1.00 0.10 1.00 0.30 1.00 0.20 0.25 0.15 1.00 10.00 1.00 0.50 1.00 1.50 1.00 1.50 0.25 10.00 1.00 0.80 0.68 1.80 0.33 1.80 0.80 1.30 0.50 10.00 0.33 2.10 1.00 1.80 0.38 10.00 1.00 10.00 0.38 I !J J J J ,)J J )J J J )J I J 1 I 1 ))].J »)))J ...1 ."j Appendix Table A-II (Continued) Percent Chinook Chinook Cover Type Cover (turbid)(clear)Coho Sockeye Chum No cover 0-5%0.15 0.01 0.00 0.18 1.00 Emergent Vegetation 0-5%0.23 0.11 0.05 0.39 1.00 6-25%0.30 0.33 0.14 0.54 1.00 26-50%0.33 0.55 0.24 0.70 1.00 51-75%0.39 0.78 0.33 0.85 1.00 76-100%0.40 1.00 0.42 ~.OO 1.00 Aquatic Vegetation 0-5%0.23 0.10 0.04 0.23 1.00 6-25%0.30 0.32 0.13 0.32 1.00 ~26-50%0.33 0.53 0.21 0.41 1.00Iw51-75%0.39 0.76 0.30 0.50 1.00-...I 76-100%0.40 0.97 0.38 0.59 1.00 Debris or Deadfall 0-5%0.15 0.05 0.08 0.21 1.00 6-25%0.20 0.17 0.24 0.29 1.00 26-50%0.20 0.28 0.39 0.37 1.00 51-75%0.20 0.39 0.55 0.45 1.00 76-100%0.20 0.50 0.70 0.53 1.00 Overhanging Riparian 0-5%0.15 0.04 0.07 0.25 1.00 Vegetation 6-25%0.20 0.13 0.20 0.34 1.00 26-50%0.20 0.21 0.33 0.44 1.00 51-75%0.20 0.30 0.46 0.54 1.00 76-100%0.20 0.38 0.59 0.63 1.00 Undercut Banks 0-5%0.23 0.11 0.12 0.25 1.00 6-25%0.30 0.33 0.34 0.34 1.00 26-50%0.33 0.55 0.56 0.44 1.00 51-75%0.39 0.78 0.78 0.54 1.00 76-100%0.40 1.00 1.00 0.63 1.00 Appendix Table A-II (Continued) Percent Chinook Chinook Cover Type Cover (turbid)(clear)Coho Sockeye Chum Large Gravel (1-3 11 )0-5%0.15 0.02 0.02 0.18 1.00 6-25%0.20 0.08 0.06 0.24 1.00 26-50%0.20 0.13 0.10 0.32 1.00 51-75%0.20 0.18 0.14 0.38 1.00 76-100%0.20 0.23 0.18 0.45 1.00 Rubble (3-5 11 )0-5%0.15 0.03 0.02 0.18 1.00 6-25%0.20 0.10 0.06 0.24 1.00 26-50%0.20 0.17 0.10 0.32 1.00 51-75%0.20 0.23 0.14 0.38 1.00>76-100%0.20 0.30 0.18 0.45 1.00, wco Cobble or Boulder 0-5%0.15 0.03 0.02 0.18 1.00 (>511 )6-25%0.20 0.11 0.06 0.24 1.00 26-50%0.20 0.18 0.10 0.32 1.00 51-75%0.20 0.25 0.14 0.38 1.00 76-100%0.20 0.32 0.18 0.45 1.00 -,~i )),J •J J J J ]]J )J I J -- - ..- also found in 1983 although discounted at the time.This difference may be due to fish reacting to high suspended solid concentrations by staying near the surface (Wallen 1951 as cited in Beauchamp et al. 1983).It also could be due to fish not being able to feed at depths where there is very little light,whereas in shallower water a greater amount of light may enable fish to feed. Coho Salmon The suitability criteria developed for coho salmon juveniles in the middle river were modified only slightly in cover suitability for use in the lower reach.The fit of the data to the composite weighting factor was not very high (r=0.32)however,which suggests that coho respond to other factors than those studied.These factors include food supply or seasonal movements. Sockeye Salmon Since sockeye normally rear in lakes (Morrow 1980),it is not surprlslng that velocity is one of the most important variables affecting their distribution.In both the lower and middle Susitna river,no sockeye were captured in cells with velocities greater than 1.2 ft/sec.The highest catches of sockeye in the lower river were made at Beaver Dam Slough,which is a backwater site with minimal velocity. Instream cover also has an effect on juvenile sockeye salmon distri- bution and it appears they use turbidity as cover (Section 3.2.4).In lakes which are turbid due to glacial input,however,production of sockeye smolts on an area basis is much smaller than that of clear lakes (Lloyd 1985).Deep water in the clear lakes would provide cover while in the Susitna,depths of 10 feet or more are infrequently found,and therefore turbidity would be used as cover.Cover type suitabilities were somewhat different in the lower reach than in the middle reach, perhaps due to differences in the primary or secondary cover type within the categories between the two reaches. Chum Salmon Chum salmon,in contrast to the other species,did not show any positive response to the presence of cover.The response shown,which is a negative one,is probably partly a function·of gear efficiency.They did respond to velocity and depth,however.The lack of relationship with cover may partly be a function of schooling behavior which reduces the need for cover.It is also possible that since chum fry rear in fresh water for only a short period,they usually are searching for food instead of hiding in cover. The reason for the heavi er use of sha 11 ower depths by chum juveni 1es found in both years not known.It could be due to a use of shallow depths and low velocities in side channels where some of the suspended solids may settle out.Perhaps these areas also are somewhat warmer than adjacent areas because the sunl ight strikes the substrate and is absorbed heating the water above. A-39 - LITERATURE CITED Beauchamp,D.A.,M.F.Sheperd,and G.B.Pauley.1983.Species pro- files:Life histories and environmental requirements (Pacific Northwest)chinook salmon.U.S.Department of the Interior,Fish and Wildlife Service.FWS/OBS-83/1. Bovee,K.D.1982.A guide to stream habitat analysis using the in- stream flow incremental methodology.Instream Flow Information Paper.No.12.U.S.Fish and Wildlife Service.FWS/035-82/26. E.Woody Trihey and Associates (EWT&A)and Woodward-Clyde Consultants. 1985.Instream flow relationships report.Vol.I.Working Draft. Al aska Power Authority Susi tna Hydroel ectri c Project.Report for Harza-Ebasco Susitna Joint Venture,Anchorage,AK.1 vol. Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt, S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.Y:-1984.Resi- dent and juvenil e anadromous fi sh investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage,Alaska. Lloyd,D.S.1985.Turbidity in freshwater habitats of Alaska.A review of published and unpublished literature relevant to the use of turbidity as a water quality standard.Alaska Department of Fish and Game,Habitat Division.Report No.85-1.l.luneau,Alaska. Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L. Crawford,and P.M.Suchanek,(eds:T.1984.Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.Z.Anchorage,Alaska. Morrow,J.E.1980.The freshwater fishes of Alaska.Alaska Northwest Publishing Company,Anchorage,Alaska. Nie,N.H.,C.H.Hull,J.G.Jenkins,K.Steinbrenner,and D.H.Bent. 1975.Statistical package for the social sciences.2nd ed. McGraw-Hill Book Co.,New York,USA. Suchanek,P.M.,R.P.Marshall,S.S.Hale,and D.C.Schmidt.1984. Juvenile salmon rearing suitability criteria.Part 3 in D.C. Schmidt,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.)-.1984. Resident and juvenile anadromous fish investigations (May -October 1983).Al aska Department of Fi sh and Game Susi tna Hydro Aquati c Studies.Report No.2.Anchorage,Alaska. Wallen,I.E.1951.The direct effect of turbidity on fishes.Doctoral dissertation.University of Michigan,Ann Arbor,Michigan,USA. A-40 - - - APPENDIX B MODELLED SITE TURBIDITIES~JUVENILE SALMON CATCHES~AREAS~SIDE CHANNEL FLOWS~ WEIGHTED USABLE AREAS~AND HABITAT INDICES .- , - This appendix is a compilation of data arranged into a number of graphs and tables.The first three tables (Appendix Tables B-1, B-2,and B-3) present:modelled side channel turbidities;modelled site catches and CPUE1s of juvenile salmon;and lengths of RJHAB model sites;respec- tively.Appendix Table B-4 presents modelled side channel flows as a function of mainstem discharge at 3,000 cfs increments. Next weighted usable areas and habitat indices are presented by species in the following order: Chinook Salmon Tabulation of weighted usable areas and habitat indices for 18 sites (Appendix Table B-5). Graphs of weighted usable area versus mainstem discharge for sites not presented in Section 3.3: Caswell Creek Mouth (Appendix Figure B-1) Beaver Dam Slough (Appendix Figure B-1) Hooligan Side Channel (Appendix Figure B-2) Bearbait Side Channel (Appendix Figure B-2) Last Chance Side Channel (Appendix Figure B-3) Rustic Wilderness Side Channel (Appendix Figure 8-3) Island Side Channel (Appendix Fi gure 8-4) Mainstem West Bank (Appendix Fi gure B-4) Goose 2 Side Channel (Appendix Figure 8-5) Circular Side Channel (Appendix Figure 8-5) Sauna Side Channel (Appendix Figure 8-6) Bearbait Side Channel (Appendix Figure 8-6) Sunset Side Channel (Appendix Figure B-7) Sunrise Side Channel (Appendix Figure B-7) Trapper Creek Si de Channel (Appendix Fi gure B-8) Coho Salmon Tabulation of weighted usable areas and habitat indices for three sites (Appendix Table 8-6). B-1 Chum Salmon Tabulation of weighted usable areas and habitat indices for 15 sites (Appendix Table 8-7). Graphs of weighted usable area versus mainstem discharge for sites not presented in Section 3.3:. Hooligan Side Channel Kroto Slough Head Bearbait Side Channel Island Side Channel Mainstem West Bank Goose 2 Side Channel Circular Side Channel Sauna Side Channel Sucker Side Channel Beaver Dam Side Channel Sunrise Side Channel (Appendix Figure 8-9) (Appendix Figure 8-9) (Appendix Figure 8-10) (Appendix Figure 8-10) (Appendix Figure 8-11) (Appendix Figure 8-11) (Appendix Figure 8-12) (Appendix Figure 8-12) (Appendix Figure 8-13) (Appendix Figure 8-13) (Appendix Figure 8-14) Sockeye Salmon Tabulation of weighted usable areas and habitat indices for seven sites (Appendix Table B-8). Graphs of weighted usable area versus mainstem discharge for sites not presented in Section 3.3:-Caswell Creek Mouth Beaver Dam Slough Sunrise Side Channel 8-2 (Appendix Figure B-15) (Appendix Figure 8-15) (Appendix Figure B-16) J })1 -)-----)1 1 )J )j } Appendix Table B-1.Turbidities within modelled side channels of the lower Susitna River,June through August,1984.Values within parentheses were calculated by inputting the overall mean for all the side channels during a given two week period. Site June 1-15 June 16-30 July 1-15 July 16-30 Aug 1-15 Aug 16-301 Mean West Bank Lateral Side Channels Kroto Side Channel Bear Bait Side Channel Mainstem West Bank Sauna Side Channel Trapper Side Channel Middle Side Channels Hooli9an Side Channel Last Chance Side Channel Island Side Channel Circular Side Channel Sucker Side Channel Sunrise Side Channel (64) (64) (64) 120 96 (64) (64) 55 89 26 18 394 392 (227) (227) 576 365 (227) 126 122 64 112 (369) 284 (369) 496 940 288 296 334 592 276 180 272,704 312 368 364 470 296 672 336 288 118 88 784 328 324 244 306 704 352 228 216 292 280 126 142 324 156,256 608 544 576 (209) 78,304 44,163 44,124 388 254 279 266 499 377 365 215 241 140 121 CD I W East Bank Lateral Side Channels Rustic Wilderness Side Channel Coose Side Channel Sunset Side Channel Beaver Dam Side Channel OVERALL MEAN (64) 41 (64) (64) 64 120 140 (227) 90 227 130 384 (369) 224 369 160 300 114 134 312 196 188 100 170 314 38 64,244 41,146 150 209 118 194 152 139 1 Two turbidities are given in this column for six sites because there were two sampling trips during this two week period in the Sunshine area.Turbidities were dropping rapidly in late August and so turbidities taken on the first late August trip were much higher than those taken during the second trip in late August. Appendix Table B-2.Catch and catch per cell (CPUE)of juvenile salmon within lower Susitna River sampling sites,1984.Cells have been standardized to an area of 300 ft 2. No.f.1t- cells Chinook Coho Chum boc:keye Chinook Coho Chum Soc:I,:eyl: S!te sampled catch C:<i\tch catch catch CPUE CPUE CPUE Cf"UE ---------------------------------~--_.__...-.-n.__......____._ -----.....-q ..... .......~_....._..~_....,---_.-..-._---,~......,.--~.q •••----------_. Hooligan Side Channel '77 21 0 lEI 3 0.27 0.00 1.01 0.04 Eagles Nest Side Channel 30 5 (I 0 0 0.17 0.00 0.00 0.00 I<roto Slaugh Head 56.5 4 0 1 2 0.07 0.00 0 ..02 0.04 Rolly Creek Mouth 91 53 ,39 2 87 0.58 0.4:5 0.02 0.96 Bearbait Side Channel 49.4 4 0 3 0 0.08 0.00 0.06 0.00 Last Chance Side Channel 50 0 0 1 0 0.00 0.00 0.02 0.00 Rustic Wilderness Side Channel 65 55 1 11 0 0.85 0.02 0.17 O.OQ Caswell Creel,:MCluth '74 419 245 (l 21 5.66 s.:'~l 0.00 0.28 OJ I~land Side Channel 82 39 1 74 ....0.4[1 0.01 0.90 0.02I"- +:00 Mainstem West Bank 45 7 (I I)1 0.16 I).(10 0.00 0.02 Goose 2 Side Channel 82 l4 1 30 "0.90 0.01 0.37 0.02"- Circular Side Channel 88 2B (I 114 6 0.32 0.00 1.30 0.07 Sauna Side Channel 44 3 (I 41 5 0.07 0.00 0.9,3 O.11 Sucker Side Channel 77.1 23 0 112 15 0.30 0.00 1.45 0.19 Beaver Dam Slough 83 14 67 0 101 0.17 0.81 0.00 1.22 Beaver Dam Side Channel 102 153 9 23 71 1.50 0.09 0 ..23 0.70 Sunset Side Channel 73.5 121 0 0 1 ','1.65 0.00 0.00 0.16 Sunrise Side Channel 73 120 1 43 8 1.64 0.01 0.59 0.11 Birch Creek Slaugh 'J6 2:::;'71 45 29 0.24 O.'74 0.47 0.30 Trapper Creek Side Channel 96 43 2 20 4 0.45 0.02 0.21 0.04 SUBTOTAL 1434.5 1209 437 598 369 0.84 0.30 0,42 0.26 Opportunistic sites 163.7 249 5 10 43 1 "')0.03 0.06 0.26•i.I..:. TOTAL 1598.2 1458 442 608 412 0.91 0.28 0.38 0.26 ],~t J ~..J J J .~t )I I "t I,)cJ J .- Appendix Table B-3.Lengths of RJHAB model sites in the lower Susitna River,1984 • .- ..... - - Site Hooligan Side Channel Eagle's Nest Side Channel Kroto Slough Head Rolly Creek Mouth Bearbait Side Channel Last Chance Side Channel Rustic Wilderness Side Channel Caswell Creek Mouth Island Side Channel Goose 2 Side Channel Sucker Side Channel Beaver Dam Slough Beaver Dam Side Channel Sunrise Side Channel Birch Creek Slough Trapper Creek Side Channel B-5 Length (feet) 1377 490 748 1437 496 961 1169 712 769 1030 658 436 608 1003 841 968 Appendix Table 8-4.Side channel flows at the 15 modelled side channels in the lower Susitna River as a function of mainstem discharge,1984.Flows calculated from rating curves presented in Ouane et al.(1985). HOOLIGAN S.C.KRDTD SLOUGH HEAD BEARBAIT SIDE CHANNEL LAST CHANCE S.C.RUSTIC WILDERNESS S,C. -~~~--~~--~~-~----------------------------------------~---------- lUi I tiSTEN SITE SITE SITE SITE SITE DISCHARGE AREA FLOW AREA FLOW AREA FLOW AREA FLOW hREA FLOW 12000 63400 (I 48200 0 3100 0 17500 0 48(1\J (I 15000 63400 0 4B200 0 3100 (I 175(/(1 0 480i)0 18000 65400 0 48200 0 3100 0 17500 0 4800 0 21000 63400 0 48200 C,3100 (;17500 0 31900 54 24000 79800 50 48200 0 3100 0 20000 1 49500 It! 27000 86900 72 48200 0 3100 (I 22000 t 60700 10~,~ 30000 908(1)100 48200 0 3100 0 2700iJ 5 6S1700 1)4 330(iO 9&500 135 48200 (I 2,100 0 34000 B 711800 171 36000 104BOO 178 50000 {S a 5700 .33 46500 13 83s00 2n 39000 113700 22'1'61900 74 10800 48 70000 21 89900 261 42000 122900 2BB 77500 98 14600 67 81000 .31 9700(1 7i~:.hJ 45000 1:31300 358 86800 128 I79(h)93 91000 46 104000 375 I 48000 141200 439 95100 163 21100 ('1'"94000 67 1(6000 442OJLJ I 51000 152000 531 102200 206 23BOG 166 9630u 95 1140(10 5160'1 5401;0 163000 636 106700 25~t 26400 217 98500 131 j I 74ll:.1 59c 57000 174100 '"'C'-,110200 ~14 29000 2i9 100200 li8 t1920()684(oJ;:' 60000 186800 B85 113500 381 31500 ~.54 101800 2.38 12li i'OO 'J7'i' 6301)0 200800 1032 116600 459 339(1(;44:1 103200 m 121700 b 660(;(i 213.300 lb'4 1190lj(l 547 .)6300 552 104400 408 12~'20(:h 69(;00 22bOOO 1373 120100 648 38})(1 b 105560 ~;:d:!1127(JO b 72(ii)t)239000 1570 121000 761 40iJOO b 106300 609 12~;OOO t 7~;£j0(;25090(:1785 12WH)889 4150('t;1070liO 844 1235(il~j b a =Flow estimated b =Ratino curve not available c =IFIM model rated unacceptable at this site flow d =Modelled at flow of 6 cfs for IFIM e =Modelled at flow of 5 cfs for IFIM f =These flows are approximate because they are heavily influenced by Cache Creek flow »}J i .1 ...1 ••I )t J J J J ._J I •j l 1 -))J i -})-)))))J Appendix Table B-4.Continued. iSLAND SIDE CHANNEL MAINSTEH WEST BANK GOOSE 2 SIDE CHANNEL CIRCULAR SlDE CHANNEL SAUNA SIDE CHM~i~EL ------------------------------------~-----------------------------------,--------,-- MA1NSTEH SITE SITE SITE SiTE SITE DISCHARGE AREA FLOW AREA FLOW AREA FLOW AREA FLOil AREA FLD~j 12000 3150(l <1 d 6160.3 {1 d (I i)59464 <.1 d 4209.5 <l e 15000 315(10 <i d 61603 ·0 d 0 (:59464 <1 d qLi)9:3 <l e 18000 31500 <1 d b16!)}{1 d I)0 59464 ::1 d 42093 <1 e 21000 3150(i d d 73426 19 13 (l 59464 <1 d 42093 q e 24000 .31500 <1 d B0904 1:,(I 0 594t:,4 <1 d 42093 <I e,J,_I 27000 31500 {1 d 93353 134 0 (I 59464 -<1 d 420S'3 <1 e .30000 31500 <1 d 108613 .307 9600 'I:59464 <1 d 42093 <i E\...!a TSi)()(1 31500 <'1 d 1147SB 470 21500 24 59464 <1 d 42(i'{3 <.1 e 36(100 392(1)6'1 117696 559 34300 {'~71590 27 -4209.3 (j e'-'" 59(iOO 45300 94 .120505 657 47800 41 7653'1 38 49127 21 c 42000 51000 126 1;J~-'"'~762 61400 52 80557 54 497:,8 25 [LC)J1i 4500 11 58500 166 129211 &74 720(j(f 65 85140 73 50289 2~ 48000 65500 215 1~~T 995 81400 81 92944 DC'50889 34OJI·:i~'u '1 '"I 51000 720(11)2n i36885 1123 c 87800 98 102530 129 51451 39-....a 54000 79400 342 140761 1260 [93200 liB 113323 j67 52011 H 5700(1 86700 424 144269 1404 [97100 141 125753 rjf1'S2b]1;50":'.11,} D(,O(H)93100 520 147B99 1555 c 9S'900 166 1:34218 Lo8 532q4 56 63000 9980(1 /;31 151842 1]1:;t 102000 195 143575 334-5427:,'10...· 66(iOO 106200 758 154205 1882 [i03200 226 150869 412 [55184 ;0 c c,90UfJ 1i 1900 904 156425 2{i57 [11)420(1 261 154657 502.c 56053 '7:i c 721)01:'118200 1070 c 158522 2241 [j(l4800 300 i57074 bl0 [57142 85 -[ ?51)O(J 123300 125&c 160818 24:)1 [10,51 i)(1 )42 15'1'211 733 c 61018 O'l r /,_11,.. a =Flow estimated b =Ratino curve not available c =IFIM model rated unacceptable at this site flow d =Modelled at flow of 6 cfs for lFIH e =Modelled at flow of 5 cfs for IFIM f =These flows are approximate because they are heavily influenced by Cache Creek flow Appendix Table 8-4.Continued. SUCKER 5IDE CHANNEL BEAVER DAN SIDE CHANNEL SUNSET SIDE CHANNEL SUNRISE SIDE CHANNEL fRAPPER CREEK j,L. -------------------------------------------------------~-------_.~---------------- l1ii WS lEN 8m SITE SITE SITE SITE DISCHARGE AREH FLOW AREA FLOW AREA FLOW AREA FLOW AREil FLDfi 12000 (I t)18900 d 49562 j e (I 0 73,300 C',- !T 151)00 I)0 1890(1 U 49562 1 e ,.0 72;30(:E f'.' 180i,/O I)0 189ilO q 49562 1 e (1 I)73300 14 f 21\;0\}(i (i 18900 <l 49562 1 e (i (,733()(i 16 f 2400i)(;0 18900 <1 49562 1 e (I (;73300 J8 f 270(i(,(J 0 18900 {1 49562 1 E'i}(I 73300 20 t 30000 8500 13 16900 <1 49562 1 e 0 0 733iJO 12 f 53000 14900 18 18900 (j 78488 47 (,0 73300 24 f 3",000 [6900 24 18'100 (1 89472 68 19000 [9 73300 26 f 3S'i)(i1)1940(;31 1890(1 (j 117943 96 53900 29 73300 28 f 42(il}O 23600 39 18900 <1 106320 132 78500 41 73300 30 f 4:;(i(;(1 29600 48 1890(J \1 122:3,)8 17B 9710(1 58 7760(,39 OJ I 480(Ji)37100 59 22400 7 j 35476 235 115400 79 91200 72 I SlUO!)46600 71 280{JCl 14924B 305 106co\,131100 1081iY,129d 54(1)(1 57900 86 32600 lB 165990 3'10 14tl900 139 12:5)1)0 221 S/UDO 6690t)101 .357{Ki 2~;173483 49:16('600 18i 137100 370 60000 ·71300 119 38i)00 45 188419 614 175600 233 15120(i 564 63000 7390(,139 39600 68 194419 7;::j 192(1(11)29~i 158000 683_oJ} 66(1uO 75900 161 40800 101 20:\000 925 207300 370 163100 81 \' b9(n)t)7730t)185 41500 148 206972 111'!c 2214(i!j 457\1669,)0 ~17~i c /~Ol):)7GlOO <:11 41900 213 210m 1345 c 2'29000 5;:,4 1hHOO 1151 c 75000 18300 24i)4210(,302 21586l i 6(i3 r.233300 688 173500 i 351 c a =Flow estimated b =Ratina curve not available c =IFIM model rated unacceptable at this site flow d =Modelled at flow of 6 cfs for IFIM e =Modelled at flow of 5 cfs for IFIH f =These flows are approximate because they are heavily influenced by Cache Creek flow ~I c..1 J )1 D I J J 1 ]}J '.,1 J J ::c - - Appendix Table 8-5.Weighted usable areas and habitat indices for juvenile chinook salmon in lower Susitna River model sites,1984. RQl.LY CREEK IHlIlTH .CASlIEll CREEK IIOIITII BEAVER DA~SLOUGH-----------_..-----_._---------------------_._-----------,..--------.------------------------------...------ IlAI1iSTElI mE CHINOOK CIlIIlOOK "AINS1£"SITE CHINOOK CIII.NIIDK "AINSTElI SHE CHINlJOK CI!!lI00K DISCHARGE AIlEA MilA Il.I.DISCHARGE ARU IlIIA II.I.DISCHARGE AREA wu~H.I. 12000 ..84'100 3900 0.05 12000 16200 800 0.05 12000 llbOO 1300 v.ll 15000 '~900 3900 0.05 15000 16200 800 0.'05 15000 llbOO 1300 0»11 18000 84900 3900 0.05 18000 10200 800 o.D5 IBOoo llWO 1300 O.ll 21000 B~'I00 3900 0.05 21000 110200 800 0.05 21000 moo 1300 O.ll 2~OOO 85300 3900 0.05 2~OOO 16200 800 0.05 HOOO 11900 1300 v.lI 21000 8lI300 3900 O.O~27000 16300 aoo 001)5 27000 12200 13M 0.11 30000 93200 3900 0.04 30000 110700 l100 0,07 30000 125\)0 1300 0.10 33000 99800 4100 0.04 33000 11300 11000 0.09 33000 13000 1300 0.10 36000 108900 4200 0.04 36000 18000 .2200 0.12 310000 13400 .1300 0.10 39000 121000 4300 0.04 39000 18900 ..2100 0.14 39000 13900 1400 0.10 42000 135000 4400 G003 42000 1'1800 3200 0.16 42000 14400 1500 0.10 45000 152/000 4500 0.03 45000 21000 1700 0.19 45000 15000 1BOD 0.12 48000 178500 7300 0.04 48000-2UIGO 4200 0.19-48000 15700 2100 0.13 51000 1'18110O WOO 0.07 51000 -mot 4700 0.21 51000 111JOO 2/000 0.16 54000 213000 20100 0.09 54000'moo.sm·0.22 54000 10800 3000 0.18· 57000 223200 23400 ll.10 57000 2_5700 0.23 57000 11600 1700 0.21 60000 229800 2S900 0.11 DOOOO 25500 6200 0.24 60000 19500 ·4200 0.23 63000 23500&28000 0.12:63000 2~6700 0.25 03000 19700 4600 0.23 66000 238700 30000 -0.13 D6000 27200 7200 0.26 610000 ~.4BOO 0.23 69000 241bOO~31500 0.13 69000 27900 7600 0.27 69000 21600 5ODO 0.23 .12000 243200 32900 0.13 ..12000 .28900 9000 0.28 72000 22100 5100 0.23 75000 243600 33500 0.14 75000 29700 8400 0.28 75000 22600 5200 0.23 HDOUGAII SIDE CHt1NIIEl ---.KRDTD S1.1lU6H HEAl IlEARBAIT SIDE CIlAIIlEi.-_..._----_._------------------------------------......------- IIAINSTElI sm CHIlIIIOK CHIIIlIOt:lIAINSTEJr SITE CHIIlIllIl CHIIIIOK ItAI~STElI SITE CHI!IIOK CHINIlOl DISCHARGE AIlEA MIlA H.I.DIS£llAli6E .AREA 1IIl~K.I.DISCHARGE AREA IIUA H.I. 12000 63400 500 0.01 12000 49200 100 .00 12000 3100 20 0.01 15000 63400 500 0.01 15000 4820t 100 .00 15000 3100 20 0.01 18000 bJ400 500 0.01 18000 48200 100 .00 18000 3100 20 0.01 21000 .bJ400 500 0.~1 21000 48200 100 .00 21000 3100 20 0.01 24000 79800 7600 -0.10 24000 48200 100 .00 2+000 3100 20 GoOl 21000 86900 7200 0.09 27000 48200 100 ,00 27000 3100 20 0.01 30000·90800 6700 0.07 30000 48200 100 .00 30000 3100 20 0.01 33000 moo 6100 0.06 33000 48200 100 .00 33000 3100 20 0.01 36000 104800 5500 0.05 3hOOO 50000 2000 o.e4 30000 5700 200 0.04 39000 113700 4900 0.04 39000 67900 4800 0.07 39000 10800 350 0.03 42000 122900 4200 0.03 42000 77500 6200 0.09 42000 14000 530 0.04 45000 131300 3600 0.03 45000 86800 7300 0.08 45000 17900 6S0 0.04 48000 141200 2900 0.02 48000 95100 BUIO 0.09 48000 21100 720 0.03 51000 152000 2200 0.01 51000 102200 7900 0.08 51000 23800 790 0.03 54000 1&3000 2000 0.01 54000 106700 6900 O.Ob 54000 26+00 aoo 0.03 51000 174100 2000 0.01 moo 110200 6000 0.05 57000 29000 750 0.03 60000 186800 1900 0.01 60000 113500 5100 0.04 60000 31500 700 0.1)2 6300(1 200800 IBOO 0.01 63000 116600 4300 a.04 63000 moo b50 0.02 610000 2lJ300 1800 0.01 06000 moDO 3400 0.03 66000 36301)610 0.02 69000 220000 1800 0.01 69000 120100 2900 0.02 69000 38300 590 0.02 72000 139000 1800 0.01 72000 121000 2500 0.02 72000 .liOOO 570 0.01 75000 250900 1800 0.01 15000 121400 22<tO 0.02 75000 41500 5bO 0.01 8-9 Appendix Table 8-5.Continued. LAST CIiA~CE S.C.~USTIC ~ILDERNESS S.C.15L~HD SIDE CH"NNEL----------------------------------------------------------------------------------------------------------- IlAIIISTEII SITE CHlNilIlk CHINOOK ~AIMSTEM sm CflINO[l(CHINiJOK ~AIIiSTEII sm CHI/JOOK CHINOOK DISCHliR&E "RE"~iIli Ii.I.DISCHARGE ARE"~u"n.1.DISCHAR6E "REA liUH H.i. 12600 moo 110 O.vl 12(1)0 4600 30 •O.iil 120W 315M 400 0.01 15000 ·mol)110 0.u1 15000 4601)30 ii.Ol 15(100 31500 400 0.01 18000 '17500 110 0.01 180M 4800 30 ii.Ol 16000 315M 400 0.01 21000 1751)0 110 0.01 21000 31901}4800 ii.15 21000 31500 400 0.01 24000 200M 1200 0.011 24000 49500 5100 O.lf)241)00 31500 400 0.01 2700(1 22000 mo O.Ob 27000 b0700 4300 0.07 moo 31500 4VO 0.01 30000 27000 1370 0.05 30000 moo 3700 0.1l5 30000 31500 400 0.01 33000 34000 1400 0.04 33000 76800 3000 D.04 33000 31500 400 0.01 3bOOO 46500 1420 0.03 36000 83300 2400 0.03 3bOOO 39M 3500 0.09 39000 70000 1440 0.02 39000 89900 1900 0.02 39000 45300 4800 0.11 42000 81&00 1470 0.02 42000 97000 1500 0.02 42000 51000 4100 0.08 45000 91000 1500 0.02 45000 104000 1260 0.01 45000 58500 3400 0.011 48000 94000 1610 11.02 48000 10'1000 900 o.vl 48000 &5500 2900 0.04 51000 96300 2050 0.02 51000 114000 100 0.01 51000 72000 2400 0.03 54000 98500 2560 0.03 54000 117400 SOO .00 54000 79400 2100 0.03 57000 1*200 2620 0.03 51000 119200 500 .00 57000 811700 18GO 0.02 60000 101800 2540 0.02 60000 120700 61)0 .00 60000 93100 .1700 0~02 63000 103200 mo 0.02 63000 121700 61)0 .00 63000 99800 1800 0~02 66000 104400 2350 0.02 116000 122200 &00 .00 66000 106200 2100 0.02 69000 105500 2240 0.02 119000 122700 700 0.01 69000 111900 2400 0.02 72000 1l>l.300 2100 0.02 12000 123000 700 0.01 72000 110200 2600 0.02 75000 107000 1900 0.02 75000 123500 BOO 0.01 75000 123300 2700 0.02 "AIIISTE"\lEST BAIIX SOOSE 2 SIDE CHANNa CIRCULAR SIDE CHAIIIl£L------------------------------------------------------------------ "AINsm SHE CIiIIfOOK CIIlllooK "AI/lSTElI SHE CHINOOK CIHNlIIIk "Al/lSTElI SHE CHllIOIIK CHtHOllK DISCHARGE AREA IIIIA H.I.DISCHARIiE AREA 1iU"H.I.DtSCHARSE AREA IIIIA H•.I. 12000 61603 1082 0.02 12000 0 0 0.00 12000 59464 i47 O~OI 15000 61603 1082 0.02 15000 I)0 0.00 15000 5941>4 747 -0.01' lBOoo 61603 1082 0.112 18000 0 0 0.00 18000 59464 747 0~01 21000 734211 10041 0.14 21000 0 0 0.00 21000 594&4 747 0.01 24000 80904 8325 0.10 24000 0 0 0.00 24000 59%4 747 0.01 27000 93353 5224 0.0iI 21000 0 I)0.00 27000 59464 747 0.01 30000 108613 4045 0.04-30000 9600 1500 0.16 30000 594&4 747 0.01 33000 11mB 3959 0.03 33000 21500 2900 0.13 33000 594&4 747 0.01 311000 117696·3861 0.03 36000 34300 4000 0.12 36000 71590 8117 0.12 39000 120505 3175 0.01 39000 47000 5100 0.11 39000 7&534 B404 0.11 42000 123397 3855 0.03 42000 &1400 &100 0.10 42000 80557 0013 O.HI 45000 129211 4113 0.03 45000 72000 6900 0.10 45000 85140 7472 0.09 48000 133649 4630 0.03 48000 81400 7000 0.09 48000 n944 7077 0.08 51000 136885 5080 0.04 51000 8711oo 6700 0.li8 51000 102530 b9911 0.07 54000 140761 5554 G.a4 54000 moo 6000 0.06 54000 1U323 119'19 0.06 57000 1442&9 6211 0.04 57000 97100 41100 0.05 57000 12575,6634 0.05 &001lQ 147899 &728 0.05 110000 99900 3100 0.03 bOOOO 134218 651&0.05 63000 1511142 7i}92 0.05 &3006 102000 2700 0.03 63000 143575 6906 0.05 6/,000 1542il5 7598 0.05 "MOO 103200 2400 0.02 06000 150869 7926 (1.05 69000 1511425 7913 O.{IS 69000 104200 21M '),02 69000 154&57 8561 0.06 nooo 158522 8078 0.05 72000 104800 1800 0.02 72000 157074 8840 0.0& 75000 1110818 6438 v.G5 75000 105100 11100 0.02 75000 15921l 81154 0.06 8-10 - - ,..,.. Appendix Table 8-5.Continued. ,.,... SAlJIU\SIDE CHANIl£L SUCKER SInE CHAHIl£l BEAYER vA"SIDE CHllMIIEl------------------------_...._..-----------------....-----.....----------------------_.._--...----------- MIM5TEN SITE CHINOOK CHIllooK IbUNSTEN SITE CHINOOK CHINOOK IIAINSTEN SITE CHINOOK CHIIIOOK r-o.DISCHARGE AREA IIlJ/l H.I.DISCHAII6E AIlO IIUA H.I.DISCHARGE MEA IlIJA H.1. 12000 42093 Ib5 .00 12000 0 0 0.00 12000 18900 50 .00 15000 >~;~2093 165 .00 15000 0 0 0:00 1~00 18900 50 .00 I BOllO . 42093 Ib5 .00 18000 0 0 0.00 18000 18900 50 .00 21000 42093 165 .00 2101»0 0 0.00 21000 18900 50 .00 24000 42093 Ib5 .00 24000 0 0 0.00 24000 IB9flO 50 .00 :nvoo 42093 165 .00 moo 0 0 ERR 27000 IB900 50 .00 30000 41093 1&5 .00 ZOOOO 8500 lObO 0.12 3ססoo 18900 50 .00 33000-42093 165 .00 33<lOO 14900 1600 0.11 33000 18900 50 .00,...., 3&000 42093 1&5 .00 3l:IOOO 1&900 1570 0.09 3&000 189flO 50 .00 3'1000 49127 5759 0.12 3~19400 1510 0.08 39000 18900 50 .00 42000 49758 5740 0.12 42000 23600 1450 0.06 42000 18900 50 .00 45000 50289 5503 0.11 ~.29600 1550 US"45000 IB'IOO 50 .00 4Bm S08119 4980 0.10 -4IIIlIf.37100 20700 .0.011: '. 22400 1120p-·48000 0.04 51000 51451 4<'10 0.09 ""1M ·46600.2940·0.0ii·.'StOOO:-2B600 2310 0.08 54000 52011 4046 0.08 .~S790f .4230 0.07 54000 moo·3560 0.1t 57000 521171l 3MS 0.07 S7OIO 1lIl900 4680·0.07 .'57000 35700 3840 0.11-60000 53294 3365 .0.06 600tll1 moo 4-490 0.011:....&0000 18000 3570 0.0'/ 63000 54275 .3116 ..0••0&1130OO 73900 4230 0.06<63000..39&00 3060 0.08 &6000 5518.-2947 o.~6&000 159flO 3940 0.05 &&000 40800 2510 '0.0& &9000 .5&053 ·.275T O.lr.i 69000 moo 1610 0.05 69000.''41SOt 2260 0;05 72000 ,5m2 .~-2'7B 0.05 72000 78100 3270 O.O~72000 419flO 2100 0.05 '75000 !1018 2714 O.O~..75000 78300 3010 0;04 1sooa'j2100 '.200t 0.05,. ...".-' SIIIISET SID£CYAilm SUIIlISE SlllECllAllllEl TIbVftircRm s.c; ,-----------_..__....-.---'-.-~......._------------ flAINSTU SITE C1UHIllllC .CH1lIOOK --IIAIIISIEI SITE CHINOOK CHINOOK IIAIIISTEIf'..SHE'CIlINlllJl('CHUIODt: DISCIIARGE AREA 11IM 8;1.DISCIlAIl6E AREA IlUA II.I~DIDR6£AREA 11IM H.I. 12000 495h2 5bi 0.01 l2GOO 0 0 0;00 12000 moo HOO 0.02 15000 495&2 568 0.01 lseot 0 0 0.00 15000'73300 1100 0.02.-18000 495&2 568 0.01 1BOOO ..0 0 0.00 ,IBOOO moo 1100 0.02 21000 49562 5bi 0.01 2100e 0 0 0.00 21000 73300 1100 0.02 24000 495&2 568 0.01 2400t 11 0 0.00 24000 n300 1100 0.02 27000 495&2 ..Shit MI 27000 0 0 0.00 27000 73300 1100 0.02 ~30000 495112 568·0))1 30000 II 0 0.00 30000'73300 1100 0.02 33000 78488 4378 0.0&31000 11 0 0.00 33000 moo 1100 0.02 3&000 89472 4420 0.05 36000 19000 ..610 0.03 3&000 moo 1100 0.02 39000 97943 4&30 0.05 39000 53900 3250 0.011 39000 moo 1100 0.02....,42000 10&320 4984 0.05 4200G 78S00 56110 0.07 42000 73300 HOO 0.02 45000 122338 5U&0.04 45000 .97100 .6090 0.0&45000 moo 9300 0.12 48000 13547&584b 0.04 41lOOO 115400 4270 0.04 48000 moo 9000 0.10 51000 149248 581>8 0.04 51000 131100 3820 0.03 51000 108100 7500 0.07 54000 165990 5768 0.03 54000 1-46900 3540 0.02 54000 i23;100 5&00 0.05 "*"57000 173483 5487 0.03 57000 160&00 3250 0.02 57000 137700 2900 0.02 60000 188419 5931 0.03 bOOOt l75bOO 3180 0.02 60000 151200 1300 0.01 63000 194419 6000 0.03 &3000 192000 3460 0.02 &3000 15800.0 1330 0.01 b6000 203000 6231 0.03 6&000 207300 3700 0.02 66000 lb3100 1360 0.01 ~/,9000 20&972 6263 0.03 69000 221400 4080 0.02 69000 166900 1390 0.01 12000 210128 6157 0.03 72000 229000 4190 0.02 72000 170700 1400 0.01 moo 2158&1 584S 0.03 1'5000 233300 4210 0.02 75000 173500 1400 0.01 - .- 8-11 - CHINOOK WUA CASWELL CREEK MOUTH 10 ~ 9 ,'" ,,;p-'" ......8::",.., a-7n....., l5.-.6~-8c: ~~5M4- 0~.3 <:) ~2 ~ 0 10 59 70 -. BEAVER DAM SLOUGH 6 ~ ............-----:;5-a- ".....,-,ii 4 c: ~UI ~&~ ~.3 0~ :I: 8 ~ ~2 ------~' 10 30 50 70 ~housands~MAINSTD4 DISC ARGE AT UNSHINE-(cfs) Appendix Figure B-1.Weighted usable area for juvenile chinook salmon at Caswell Creek and Beaver Dam tributary study sites as a function of mainstem discharge. B-12 - - - "... 0.9 .......0.8..;-d'0.7It....., ~0.6~~c~5:0.5~~~0.4 a~0.3"5~0.2 0.1 0 -_._--- 10 BEARBAlT SIDE CHANNEL 3Q SO 70 (Thousands)_ MAlNSTEM DISCl-t~GE AT SUNSHINE (cfs) Appendix Figure 8-2.Weighted usable area for juvenile chinook salmon at the Hooligan and Bearbait Side Channel study sites as a f~nction of mainstem discharge. ~8-13 2.B 2.6 2.4- ,... 2.2i 0-2 II) '-' ~1.8 ~-S 1.6c: ~:Ul 1".4-~~1.2::£. 0~O.BG (;j 0.6~ 0.4- 0.2 0 10 CHINOOK WUA LAST CHANCE SIDE CHANNEL Br••clt.d t .30 50 70 (Thousands) MAINSTEM DISCH~GE:!<T SUNSHINE:(cts) RUSTIC WILDERNESS SIDE CHANNEL 90 - .... 6 5,... ..;-cT I) '-'4- ~~-Sc:~g 3 Bre8ched~~t::£. 0 2~ () ~ - o ----- 10 30 50 70 .(Thousands). MAlNSTEM DISCli!'RGE;!'oT SUNSHINE:(cts) ""'" Appendix Figure B-3.Weighted usable area for juvenile chinook salmon at last Chance and Rustic Wilderness Side Channel study sites as a function of mainstem discharge. B-14 - - Appendix Figure B-4.Weighted usable area for juvenile chinook salmon at the Island Channel and Mainstem West Bank study sites as a function of mainstem discharge. 8-15 '""'" ~ CHINOOK WUA GOOSE 2 SIDE CHANNELe ""'" 7 ::;--6e- li..... ~5 c ~i 4 ~~~.:5 .-0~:r ~2 ~---- ~ 0 10 30 59 70 ".,., CIRCULAR SLOE CHANNEL 10 IlI'ucfMtd 9 ......e...-U 7II.....-~6~-8c ~i 5~.~~ :£.4- 0 e!.:5:r Cl -iii~2 1 0 10 30 50 70 ~uSandS~MAINSTEM DISC GE AT UNSHINE (cfs)- Appendix Figure 8-5.Weighted usable area for juvenile chinook salmon at the Goose 2 and Circular Side Channel study sites as a function of mainstem discharge. 8-16 70 ........ ....................... BEAVER DAM SIDE CHANNEL4-.------..--:=.:..::..::.:.:....:....:..--------------, 2 3 1 o ---------------- 10 30 50 (Thousands) MAlNSTEW DISC~GE:AT SUNSHINE (cfs) 2.5 1.5 3.5 0.5 ,... I -- - Apoendix Figure 6-6.Weighted usable area for juvenile chinook salmon at the Sauna and Beaver Dam Side Channel study sites as a function of mainstem discharge. ~B-17 30 50 70 (Thousands). MA1NSlEM DISCHARGE AT SUNSHINE (cfs) CHINOOK WUA SUNSET SIDE CHANNEL 7 """ 6 ...... ~..,. 1:1 5II...... ~~-a 4 Breaclledct~i!~3~ a ..- ~2J:~~ ~ 0 10 30 59 70 1"'''' ~ SUNRISE SIDE CHANNEL 7 6 ......Breeched~t1:1 5 ~iII..... ~~-a 4c ~i ~~~3:£ a~2 C) iii~ o-l----_---.....-..-.I!..---r---r-----r-----,r------i 10 Appendix Fiqure 8-7.Weighted usable area for juvenile chinook salmon at the Sunset and Sunrise Side Channel study sites as a function of mainstem discharge. 8-18 )1 -1 ))1 1 1 .j -1 l } 10 9 .-;.8....-a 7I/) '-" ~6 c ~i 5~~:>t:,4 co I 0I~......31.0 J:<:>w ~2 1 0 10 CHINOOK WUA TRAPPER CREEK SIDE CHANNEL Projected WUA ..: (Head barely overtopped) ~---- 30 50 70 (Thousands) MAINSTEM DISCHARGE AT SUNSHINE (efs) Appendix Figure 6-8.Weighted usable area for juvenile chinook salmon at the Trapper Creek Side Channel study site as a function of mainstem discharge. Appendix Table B-6.Weighted usable areas and habitat indices for juvenile coho salmon in lower Susitna River model sites,1984. - - ROLLY CREEK IIOOTH CASliELL CREEk rtOUTH BEAVER DA~SL IlUSH ..-.-------------------...-_..._----_..----_______a ____________________________-_..--------------------------_..----- ~AIIiSTE"SITE COHO COHO ""IIISTE"SHE COHO COHO "AIIISTE"sm COHO ~OHD OlSCHAA6E AREA lIUA H.I.~ISCHARSE ARE ..lIUA H.L DISCHARGE AREA lIUA H.I. 12000 84900 7900 0.09 12090 lb200 1350 i).os 12000 11600 1700 1).15 15000 84M:1'100 0.09 1500t}lb20{)1350 (i.03 15(IC'{1 !tbOO 170')0.15 18000 .84900 mo 0.09 18(/1)0 Ib200 1350 0.08 180(;('llbOO 1700 0.15 21000 84900 7900 O.M 21000 16200 1350 o.oe moo woo 1700 0.15 24000 85300 7900 0.j)9 24000 Ib200 135(i 0.08 24000 moo 170C'1).14 27000 88300 /700 0.09 27000 16300 1500 u.O'1 27000 12200 mil 0.14 30000 moo 7500 .O.OB 30000 16700 1700 0.10 10000 moo 1700 0.14 33000 998(1)mo 0.07 33000 moo 2000 (I 1"1 33000 13000 1700 0.13... 36000 109900 6900 0.00 30000 18000 2300 0.13 36000 13400 17n0 0.13 39000 121000 .1I~00 0.05 390liO 18900 2500 0.13 39000 moo 1700 0.12 42000 135000 .5900 0.04 42000 1'1800 2800 0.14 ~2000 1~400 1610 0.12 45000 152600 5500 Q.04 45000 21000 3000 0.14 4liOOO 15000 16liO 0.1l 48000 178500 51100 0.03 48000 21800 3200 0.15 48000 15100 1610 0.10 51000 198800 1300 0.04 51000 22700 340G 0.15 510GO 16300 1s.40 0.09 54000 213000 9200 0.04 5-4000 moo 3600 0.15 54000 16BOO 1480 0.09 57000 223200 10100 0.05 51000 246GO JBOG 0.15 510GO moo 'IUO 0.08 IIGOOO 229900 1070G G.OS 60000 25500 ~ooo 0.16 bOOOO IBliOO Hao O.OB 63000 23'.'i001)11200 0.05 moo 20300 ~300 0.16 63000 1910G 1540 0.G8 66000 238100 11700 G.05 60000 moo 4400 0.10 60000 20BGO 1630 0.08. 69000 24t1100 120W 0.05 09000 279GO ~700 0.17 69GOO 21600 1740 0.08 72000 m200 12300 G.05 72000 28900 4'100 0.J7 72000 22100 1780 0.08 75000 243000 12500 G.OS 75000 29100 5100 0.17 75000 22600 1B1'>0.li8 Apoendix Table B-7.Weighted usable areas and habitat indices for juvenile chum salmon in lower Susitna River model sites,1984. HOOLlIlAll SIll£CIIIlNIIEL KRIITO SlIllf6ll HEAD BEARBAIT SIIIE CHANNEL---------_._----------_...---_..-------------------------------------------------- MIHSTEII SITE ..CIlUIl &HUll IlIlINSTEII SITE CHUII CIIUII ~AINSTEIl sm (HUll COOl! IIISCHARGE A~"lIUA H.I.IIISCHARGE AREA IIUA H.r.IIISCI!/lRBE AREA WUA H.1. 12000 63400 28500 0.45 12000 48200 39600 0.92 12000 31\)0 13GO 0.42 15000 63400 28500 0.45 15000 48200 39600 0.82 15000 3100 1300 0.42 18000 63400 28500 0.45 18000 48200 39600 0.82 18000 3100 1300 0.42 21000 63400 28500 0.45 21000 48200 39~0 0.82 21000 3100 13GO 0.42 24000 79900 47'100 0.60 24000 48200 39600 o.B2 24000 3100 1300 0.42 27000 i6900 46700 0.54 27000 48200 39600 0.82 27000 3100 1300 0.42 3ססoo 901100 44000 0.48 30000 ~82OO 39000 0.82 3ססoo 3100 1300 0.42 33000 96500 moo Q.43 l3OllO 48200 39600 o.a2 33000 3100 1300 0.42 36000 104l1oo 38400 0.31 -_.-3bOOO.5000&39600 0.79 36000 5700 I~OO 0.25 l'iOoo Imoo 34700 0.31 ~19000 61900 42000 0.62 39000 108GO 1900 0.19 42000 122'100 30300 0.25 -,;:,'42000 11500 44500 0.51 42000 ·14600 2600 0.18 45000 111300'16100 o.~·-·:r:-.B6IIOO "100 .0.53:45000 17900 330Ct 0.18 48000 141m 21900-<t.lo·..-4801»95100 4160t 0.50 48000 moo 4100-0.19 ~1000 152000 ill900 0.12 .- 51000 102200 4lJ5Oec c-Q.45 51000 23800 5300 .0.22 54000 163000 19100'.0.11 54000 106100·-.42300 0.40 54000 2&400 5700 0.22 57000 174100 .17600.:0.10 57000 110200 38lOO 0.35 57000 29000 5500 0.19 6ססoo 18llBOO:.11.200.0.09 60000 113500 ·34400 0.10 60000:31500.5100 0.1& 03000 200800 10900 0.08 63000 116600 29100 0.21 .6lOOO 33900 4700 0.14 66000 113300 16700 0.09 66000 H'lOOO 24100 0.20 66000 l6300 «00 0.12 69000 226000 IMoo 0.07 6'1000 120100 19800 0.16 69000 38300 4200 0.11 72000 239000 16100 0.07 .12000 121000 17800 0.15 72000 4000t ,4100 0.10 75000 250900 ·l511oo 0.011 15000 121400 15200 0.13 1liOOO'41500 4000 tl.l0 8-20 - -- Appendix Tab1e 8-7.Continued: LAST CIIAIlCE S.t.-.RUSTlC 1I1LDERNESS S.C.·ISlAilD SIDE CIIAlfIIEl.-----_.---~----------.---------------______________...__..:.._-.a.____ IVIIIISTU SITE OWl!.CHIlli IIAIIISTEJI SITE CHUII ClIIII IIAINSTEJI SITE CHill .CIIJII DISCIlARSE AREA lRIA H.I.DISClfAR6E AREA IIIlA H.I.IlISCHAR6E AREA MIlA H.I. 12000 17500 11500 o.oll 12000 4800 3600 0.75 12000 .moo 19300 0./11 15000 17500 11500 0.66 15000 4800 3600 0.75 15000 31500 19300 0.111 IBOOO .17500 11500 0.66 laooo 4BOO 3600 0.75 18000 31500 1'1300 0.61 21000 11500 1/500 0.66 21000 31900 30800.0.97 21000 31500 19300 0.61 24000 20000 11500 0.58 ·24000 49500 32500 D.66 24000 31500 moo 0.61 27000 22000 moo 0.52 27000 60700 27600 0.45 27000 31500.19300 0.61 30000 27000 11500 0.43 30000 69700 22700 0.33 30000 31500 19300 G.61 33000 34000 moo 0.34 33000 70800 18100 0.24 33000 31500 19300 0.61 36000 46500 11500 0.25 36000 83300 13700 0.16 36000 39200 28100 0.72 39000 70000 1/500 0.16 39000 89900 10600 0.12 3~00 45300 moo 0.lI4 42000 81000 11500 0.14 42000 97000 8000 0.09 42000 51000 25800'0.51 45000 91000 11500 0.13 45000 104000 7400 0.07 45000 sa500 22700 0.39 48000 94000 moo 0.12 48000 109000 saoo 0.05 49000 65500 /9700 0.30 51000 96300 15100 O.lb 51000 114000 4200 0.04 51000 72000 17400 0.24 54000 'tlI500 20200 0~21 54000 117400 3300·0.03 54000 moo 15100 0.19 57000 100200 moo 0.19 57000 119200 3000 0.03 57000 86700 l3200 0.15 60000 101800 18000 0.18 60000 120700 3000 0.02 60000 moo 12400 0.1l moo 103200 16200 0.16 63000 121700 3000 0.02 63000 99800 12700 0.13 66000 104400 13600.0.13 66000 122200 3000 0.02 6bOOO 106200 13000 0.12 69000 105500 10500 0.10 69000 122700 30UO 0-02 69000 111900 moo 0.12 72000 106300 8800 0.08 72000 123000 3000 0.02 moo JJB20D moo 0.12 75000 107000 7600 0.07 75000 123500 3000 0.02 75000 123300 13600 0.11 "AINSTEII iiEST BAlik 600SE 2 SIDE CHAIIIlEL CIRCULAR SiDE CHANNEL -------_..._-------------------------------------------------------_._--......--------------_._------------- ~AINSTEII SITE (Ill!"CHIlli IIAINSTa SITE l:HUII CHIJIl ~AINSTElI SITE GJiU"CHUII DISCHARBE AREA lIUA H.1.DISCHARGE AREA ~UA H.L DISCHARGE AREA MUA H.I. 12000 6160~47090 ,).76 J2000 0 0 Uo 12000 59464 411109 0.78 150M bl~"4mo 0.76 15000 0 0 0.00 15000 59404 46109 0.78 18000 II 160l 470%iJ.76 18000 0 0 0.00 181)00 59464 46109 0,78 2100\)73426 53955 0.73 21000 0 0 0.00 21000 59464 46109 0.78 24000 80904 43289·0.54 24000 0 0 0.00 24000 59464 46109 0.79 27000 93353 316011 0.34 27000 0 0 (;.00 27000 59464 46109 0.78 30000 108613 27151 0.25 30000 9600 4900 0.51 30000 59464 411109 0.78 33000 114738 23420 0.20 33000 21500 11000 0.51 33000 59464 46109 0.78 36000 117696 m82 0.19 36000 34300 17400 0.51 36000 71590 44495 0.62 39000 .120505 21096 0.18 39000 47800 25500 0.53 39000 76534 4%06 0.58 42000 J23397 2/218 0.17 42000 moo 31800 0.52 42000 .·80557 42269 0.52 45000 129211 22389 0.17 45000 72000 37900 0.53 45000 85140 42176 0.50 48000 133649 26710 0.20 4BOO0 81400 moo 0.51 48000 92944 43074 0.46 51000 136885 27661 0.20 51000 87800 42600 0.49 51000 102530 45026 0.44 54000 140761 30382 ·0.22 54000 93200 40700 0.44 54000 113323 50073 0.44 57000 .144269 :U815 0.22 57000 97100 33400 O.M 57000 115m 50248 0.40 60000 147B9'1 33950 0.23 60000 99900 24000 0.24 60000 134218 46305 0.34 63000 151842 35'153 0.24 63000 102000 18600 0.18 63000 143575 49339 0.34 66000 154205 364.89 0.24 66000 103200 13800 0.13 66000 150869 49565 0.33 09000 1511425 30211 0.23 119000 104200 10400 0.10 119000 154657 50346 0.33 72000 158522 37029 0.23 72000 104800 Il3OO 0.08 72000 157074 48491 0.31 75000 160818 3680'/0.23 75000 105100 7400 0.07 75000 159211 .46797 0.29 \ 8-21 Appendix Table B-7.Continued. SAUIlA SIDE CIWllIEl SUO:ER SIDE 0IAIIIlEI.B£AYER llM 51*CJlAtlIIEL ----------------------------------------------------------------------- MIIISTEII SITE CIlIlIl CIilII IlAINSm SITE CHUIl C!lIJII lIAINSTEII SITE CHUIl CHIllI DISCHAR6E AREA ~H.I.DlSCHAfl6E AAEjl IItIA H.1.DISCHAR6E AREA IiIIA H.I. 12000 420'13 31754 0.75 12000 0 0 0.00 12000 18900 11900 0.113 15000 42093 31754 0.75 15000 0 0 0.00 15000 1890ct .11900 0.113 18000 42093 31154 0.75 18000 0 0 0.00 18000 18900 11900 0.113 21000 420'13 31754 0.75 21000 0 0 0.00 21000 18900 11900 0.113 24000 420'13 31154 0.15 24000 0 0 0.00 24000 18900 11900 0.113 27000 42093 31754 0.15 27000 0 0 ERR 27000 18900 11900 0.113 30000 42093 3175'0.75 30000 8500 7300 0;86 30000 18900 11900 0.03 33000 .42093 31754 0.15 33000 14900 11800 0.79 33000 18900 11900 0.113 30000 42093 31754 0.75 311000 111900 12700 0.15 311000 18900 11900 0.113 39000 49127 27301 0.511 39000 19400 13200 0.08 39000 18900 11900 0.03 42000 49758 20413 0.53 42000 231100 13400 0.57 42000 18900 11900 0.113 45000 50289 25204 0.50 45000 29boo 14300 0.48 45000 18900 11900 0.03 48000 5OBll9 231170 0.47 48000 moo 19900 0.54 48000 22400 13200 0.59 51000 51451 22565 0.44 51000 4&000 27700 0.59 51000 28000 15700 0.5b 54000 52011 218311 0.42 54000 57900 33100 0,58 54000 moo 17500 0.54 57000 52078 21381 0.41 57000 6&900 34#0 0.51 51000 35700 18800 0.53 00000 53294 20990 0.39 110000 71301)32900 0.411 110000 38000 18200 0.48 113000 54275 206&9 0.38 03000 73900 W800 0.42 63000 39000 Ib400 0.41 66000 55184 20938 0.38 116000 75900 28200 0.37 cbOOO 40800 14000 0.34 &9000 511053 21017 0.37 &9000 moo 25000 0.32 09000 41500 12100 0.29 72000 57142 21153 0.37 72000 78100 21800 0.28 72000 41900 11300 0.27 75000 c10t8 :3075 Q.38 moo 78300 19200 0.25 75000 42100 10700 0.25 SU/iSEl SIDE eHA~~El SUNRISE SIDE CHANNEL TRAPPER CREEK S.C. ------------...------------------------------------,..----------------------------------_..------ "AINSlEIl SiTE CHUII CHIlli IlAINSTEIl SITE CHUIl CHUIl IlAIIISTEIl SITE CHUIl eHU" DISCHARGE AREA WA H.I.DISCHAR6E AREjl lIUA H.1.DISCHARGE AREA WUA H~L 12000 49562 27135 0.55 12000 I)0 0.00 J2000 moo 45400 0.62 15000 ·mo2 27135 0.55 15000 0 0 O.Oll 1S{)00 73300 45400 0.62 lBOOO 49562 27135 0.55 18000 II 0 0.00 18000 moo 45400 O.b2 21000 49562 27135 0.55 21000 0 0 0.00 21000 ;3300 45400 0.62 24000 4951>2 27135 0.55 24000 0 0 0.00 24000 73300 45400 O.b2 27000 49502 27135 0.55 27000 0 0 0.00 27000 moo 45400 0.112 30000 495112 27135 0.55 30000 0 0 0.00 300Qtl 73300 45400 0.112 33000 78488 34059 0.43 33000 0 0 0.00 33000 73300 45400 0.62 36000 89472 34808 11.39 36000 19000 11200 0.33 36000 73300 45400 0.il2 39000 97943 37649 0.38 39000 53900 32400 0.00 3'1000 73300 45400 0.62 42000 106320 39888 0.38 42000 ·7~464110 0.5'1 42000 13300 45400 0.62 4SOOO 122338 46376 0.38 45000 97100 moo 0.51 45000 71000 44800 0.58 48000 1354711 51185 0.38 48000 115400 44500 0.39 4BOOO 91200 41200 0.45 51000 149248 52D7t o.~51000 131100 37500 o.:n 51000 108100 3%00 0.32 54000 11.59'!0 53786 0.32 54000 146900 31100 0.21 54000 123300 27500 0.22 57000 1734113 48410 0.28 57000 16~·20600 0.17 57000 .·137700 1'l5OO 0.14 00000 18841'1 50093 o,'n 60000 1151>00 25200 0.14 60000 151200 10700 0.07 03000 19441'1 .43299 0.22 1>3000 192000 25300 0.13 113000 158000 10200 0.06 66000 203000 41715 0.21 66000 207300 211200 0.13 116000 163100 10000 0.06 69000 206972 31100 0.18 69000 221400 21700 0.13 11'1000 166900 9800 o.ob 72000 210728 33481 0.16 72000 229000 28500 0.12 72000 170700 91100 0.011 75000 215861 32949 0.15 75000 233300 29000 0.12 75000 173500 9500 0.05 I B-22 ~I .... - - - - -CHUM WUA HOOUGAN SIDE CHANNEL 50 45 ..... i 40 0- "..... ~35 ~-ac ~lil .30.~a:£250~ Cl 20i>i~ 15 10 .-10 30 50 70 90 110 .~USQndS~ MAINSTEM DISC _GE AT .UNSHINE (cfs) KROTO SLOUGH HEAD 50 ~,45 ........40 ------- ~0- ".....35 ~~-a .30c ~g ~a 25:£ 0 20~~:I: Q ~15 10 ~~, 5 10 30 50 70 90 .-~hOUSQndS~. MAINSTEU DISC ARGE AT UNSHINE (cfs) Appendix Figure B-9.Weighted usable area for juvenile chum salmon at the Hooligan Side Channel and Kroto Slouqh Head study sites as a function of mainstem discharge .. B-23 CHUM WUA BEARBAIT SIDE CHANNEL 7 6 ..........-0-5II)....., ~~~4c ~a~5 3::£ ~2:I: Q l.lJ~- 0 -10 S9 70 ~ ISLAND SIDE CHANNEL 36 34 ~ 32........;30-0-281ft....., ii 26 c 24~g ~5 22 ~ ::£20 ~18~16 14 12 10 10 30 50 70 ~OUSandS~MArNSTEM DISC ARGE AT UNSHINE (cfs)~ Appendix Figure 8-10.Weighted usable area for juvenile chum salmon at Bearbait and Island Side Channel study sites as a function of mainstem discharge. 8-24 -CHUM WUA MAINSTEM WEST BANK 54-52 50 ""48-46-0-44II '-' ~42 40 c:38WO ...Jill ~2 36 ::It 34 ,-0 32~30z Cl 28i£l 3:26 24 22 20 10 30'59 70 - Appendix Figure 8-11.Weighted usable area for juvenile chum salmon at the Mainstem West Bank and Goose 2 Side Channel study sites as a function of mainstem discharge. B-25 - ""'" CHUM WUA CIRCULAR SIDE CHANNEL.55 ~ 54 53 ,....52 -~ a-51 II 50....." ~49 Breached~~c 48 t~51 ~5 47 :>t 46 """"0 45~ J:44C) iii~43 42 41 40 10 30 59 70 -SAUNA SIDE CHANNEL 35 34 ar••cmeci 33 t,...32~ cT 3~ II 30....., ~29~-8c 28~~ ~5 27 ::£.26 0 25~J:240 iii~23 22 21 20 10 30 50 70 ghousandS~ MAINSTEIvI DISC ARCE AT UNSHINE Cds) Appendix Figure B-12.Weighted usable area for juvenile chum salmon at the Circular and Sauna Side Channel study sites as a function of mainstem discharge.- B-26 .-- CHUM WUA SUCKER SIDE CHANNEL 40 35-"...., ~30 0-......, ~25 ~~c'jg 20 ,,~5 Breached ""~1S ta ~ ~ 0 10i:jj ~ S 0~ 10 30 50 70 SEAVER DAM 510£CHANNEL 20 19 16 17 ........16~15 0-14......., i1 13 12 c 11 ~i 10~~9~6 0 7~6~ ""'"Cl 5i:jj ~4 3 2 1 a 10 Breached t ....---------------".... ".................. 30 50 70 (Thousands) MAINSTEM DISCHARG~AT SUNSHINE (cts) Appendix Figure B-13.~Jeighted usable area for juvenile chum salmon at the Sucker and Beaver Dam Side Channel study sites as a function of mainstem discharge. B-27 CHUM WUA SUNRISE SIDE CHANNEL 60 I I 50,-... i cT (I) '-"40 li-.. ~-8c 'j~30 -,I I "---------~~:>t, 0 20coI~I N ::r:(Xl e" W~ 10 30 50 70 .(Thousands) MAINSTEM DISCHARGE AT SUNSHINE (cfs) a I-T "T.1 iii I I 10 Appendix Figure B-14 ..Weighted usable area for juvenile chum salmon at the Sunrise Side Channel study site as a function of mainstem discharge. )_I J J j j _J J -J t I J ))I J 1 1 J I J 1 1 ))J j ))~ J o::J I N l.O ~ppend1~T~~le ~-~.~~19~~~~y~~~le a.r~~~~n~h~~1~~t 1n91~~$,fer j~Y~n11~SockeY~s~'~9n 10 19~~r ~Y$ftn~B1Y~F ~9gel s1t~S!1984. ""..,",,'.''''-,,-,,,,''''''''''''''''''''''''''''''-''''''",','"''"","""'",'","".""" ROLLY CREEK "OUTH CASWELL CRW IIOUT~IUVER Mil SLOIJ6H--...-----_...........----------..__..._..._......-...--_.._--_...~.....------_...............................----...~_..----...-------_......_...,..-........_...........--_.. nAI1l5IEIl SUE ,SOCKEIE SGUHE "A INsTEn SUE SOCKEYE SOCKEIE itA INSIEft SITE SOCKEYE SUCKE,E ~ISCHARGE AREA NUA H.I.DISC~AR6£AREA'lIUA H.I.DISCHAR6E AREA NUA H.J. 12000 84900 10600 0.12 12000 16200 1350 0.08 12000 11600 6200 0.53 15000 84900 10600 0.12 15000 16200 1350 0.08 15000 11600 620\)O.S~ 18000 94900 10600 0.12 18000 10200 1350 0.08 18000 11600 6200 O.SJ 21000 84900 10600 0.12 21006 WOO 1350 0.08 21000 11700 6200 0.5: 24000 85300 10600 0.12 24000 16200 1600 0.10 24000 11900 mo (0.52 27000 88300 11000 0.12 27000 16300 1700 0.10 27(100 12200 64(0)O.S: 30000 moo 13400 0.14 30000 16700 1900 0.11 30000 12500 660·)&553 33000 99800 17600 0.18 33000 17300 2300 0.13 moo 13000 6700 v.52 36000 108900 22900 0.21 36000 18000 2600 0.14 36000 13400 7000 0.5: 39000 121000 28900 0.24 39000 18900 3100 0.16 39000 moo 1100 0.51 42000 135000 35500 0.26 42000 19800 3100 0.19 42000 14400 1300 1).51 45000 152600 43400 0.28 45000 21000 4300 0.20 45000 15000 1500 0.50 48000 178500 52100 0.29 48000 21800 5000 0.23 48000 15700 7700 0.49 51000 198800 64400 0;32 51000 22700 5700 0.25 51000 16300 BOOO 0.49 54000 213000 75300 0.35 54000 23700 6400 0.27 moo 16800 8200 O.H 57000 223200 B2800 0.37 57000 24600 7200 0.29 57000 17600 B600 0.49 60000 '229800 88200 0.18 60000 25500 7900 0.31 60000 18500 8900 0.48 63000 235000 93000 0.40 63000 26300 8600 0.33 63000 19700 9400 0.4S 66000 238700 moo 0.41 66000 27200 9200 0.34 66000 20800 10200 0.4'1 09000 241600 99900 0.41 69000 27900 10000 0.36 69000 21600 10800 0.5~ 72000 243200 100700 0.41 72000 28900 10600 0.37 72000 22100 11000 0.5e 75000 243600 101500 0.42 75000 29700 11400 0.38 •75000 22600 11000 0.4~ SUCHR SIDE CHIlH~EL BEAVER DAft SIDE CHANhEL SU~SET SID£CHANNEL SUNRISE SIDE CHANNEL ----------_........_---...-------_.._..._---_....----------_..._....._-_........._----_..-------.._.._-.....----------_....._-..._--------...-----_..._---------------------------...- ItAINSIEft SITE SOCKEYE SOCKEYE ItAINSlEH SITE SOCKEYE socmE "AINSTEH SITE SOCKEYE SOCKEYE "AI~STE"SITE SOCKEYE SOCKEYE DISCHARGE AREA WUA H.I.D[SCHAR6E AREA NuA H.I.DISCHARGE AREA llUA H.I.DISCHARGE AREA ~UA H.I. 12000 0 0 D.Qo moo 18900 300('0.16 12000 49562 7182 0.14 12000 0 0 0.00 15000 0 0 0.00 15000 18900 1000 0.16 15000 I 49562 7182 0.14 15000 o .0 0.00 18000 0 0 0.00 18000 18900 3000 0.16 18000 49562 7182 0.14 18000 0 0 0.00 21001i '0 0 0.00 21000 18900 3000 0.16 21000 49562 7182 0.14 21000 0 0 0.00 24000 0 0 0.00 24000 18900 1000 0.16 24000 49562 7182 0.14 24000 0 0 0.00 27000 0 0 ERR 27000 IBM 3000 0.16 27000 49562 7182 0.14 27000 0 0 0.00 30000 8500 1200 0.14 30000 18900 3000 0.10 30000 49562 71112 ··0.14 30000 0 0 0.00 33000 14900 1800 0.12 33000 18900 3000 0.16 33000 78488 6738 0.09 33000 .0 0 0.00 16000 16900 1100 0.10 36000 18900 3000 0.16 36000 89472 6493 0.07 36000 19000 400 0.02 19000 19400 1500 0.08 39000 18900 3000 0.16 39000 97943 6639 0.01 39000 53900 4700 0.09 42000 moo 1200 0.05 42000 18900 3000 0.16 42000 106320 6828 0.06 42000 78500 5800 0.01 45000 29600 1200 0.04 45000 18900 3000 0.16 45000 122338 7412 0.06 45000 '91100 51100 0.06 48000 37100 2600 0.07 48000 22400 3200 0.14 48000 135476 7529 0.06 48000 115400 3400 G003 51000 46600 4000 0.09 51000 28000 3700 0.13 51000 149248 7108 0.05 51000 131100 3200 0.02 54000 57900 5000 0.09 54000 32600 4100 o.n 54000 165990 6643 0.04 54000 146900 3100 0.02 57000 66900 5300 0.08 57000 35700 4300 0.12 57000 173483 600b 0.03 57000 160600 3000 0.02 60000 moo 5400 0.08 60000 18000 4100 0.11 60000 188419 6662 0.04 60000 175600 3000 0.02 63000 moo 5500 0.07 63000 39600 3900 0.10 61000 194419 6275 0.03 61000 192000 3100 0,02 66000 15900 5600 0.07 66000 40800 3600 0.09 66000 203000 6740 (1.03 66000 207301)31(10 0.01 69000 moo 5600 0.07 69000 41500 3200 0.08 69000 206972 6650 0.03 69000 211400 3200 0.01 72000 78100 5600 0.07 72000 moo 3000 0.07 nooo 210728 7m 0.03 72000 229000 3200 0,01 75000 78300 5600 0.07 75000 4-2100 2800 0.07 75(100 215861 7661 0.04 75000 233300 :;200 0.01 SOCKEYE WUA CASWELL CREEK MOUTH ~,14 13 , '" ,12 '".-....... 11i &10 -!. ~9 ~ ~-S 8c: ~g 7 ~5 6 """:>E.. 0 5 /!:!4J: Qw 3~ 2 0 ~ 10 30 59 70 -BEAVER DAM SLOUGH 13 ~ 12 .-. i ...-&11 .....- ~,~ ~ ~-S 10c: ~g ~~5 9:>E.. 0 /!:!8 ~J: Cl iii ~, ------ 6 10 30 50 70 ghousandS~. MAINSTEM DISC ~G_E .~T_UNSHINE (cts) Appendix Figure B-15.Weighted usable area for juvenile sockeye salmon at Caswell Creek and Beaver Dam tributary study sites as a function of mainstem discharge. 8-30 )))1 1 1 j .•~J 1 )]j I ]»1 J SOCKEYE WUA SUNRISE SIDE CHANNEL 7 ,, 6.- "'"':'.......... r:r 5 - lD '-'" ~ ~-8 4~ c ~~3~Sreached,0...-----"-------~~t:It 0 OJ I ~2 - ,:x:w Q..... ~ 1 - o-+-i i J i I I I I 10 30 50 70 .(Thousands) MAINSTEM DISCHARGE AT SUNSHINE (cis) Appendix Figure B-16.Weighted usable area for juvenile sockeye salmon at Sunrise Side Channel study site as a function of mainstem discharge. - APPENDIX C COMPARISON OF THE IFIM AND RJHAB MODELLING TECHNIQUES AT TWO SEL ECTED SITES - ..... INTRODUCTION In 1983,two techniques were used to model the effects of mainstem discharge on juvenile salmon habitat within the middle Susitna River. The Instream Flow Incremental Methodology (I FIM)(Bovee 1982)was used at seven sites (Hale et al.1984)and the RJHAB habitat model developed in Marshall et al.(1984)was used to model six other sites.Since studies of the effects of mainstem discharge on juvenile salmon habitat within the lower Susitna River were begun in 1984,it was desirable to compare these two modelling methods.Both methods were used,therefore, at the same transects withi n two sites to compare resul ts from the two techniques. METHODS Trapper Creek Side Channel (RM 91.6)and Island Side Channel (RM 63.2) were selected as sampl ing sites for this comparative study because they represent two different channel types of the lower Susitna River. Trapper Creek Side Channel is a simple straight channel.Island Side Ch~nnel is a more complex,winding channel.Further descriptions and photos of these two sites are contained in Quane et al.(1985). Descriptions of the two modelling techniques will not be presented here. Detailed descriptions of the IFIM are presented in Appendix D of this report and Bovee (1982),and summarized in Section 2.0 of this report. The original RJHAB model was first developed and described in Marshall et al.(1984)and modifications were described in Section 2.0 of this report. Both techniques entail taking depth,velocity,and cover or substrate measurements spaced at intervals across transects runni ng at ri ght angles to the channel.Hydraulic models which have been developed for use in the IFIM include the IFG-2 model which is based on open channel flow theory and one set of field data and the IFG-4 model which is based more strongly on fi el d data as three sets of fiel d measurements are recommended (Milhous et al.1981).Fewer measurements are taken for each RJHAB field data set than for the IFIM models but up to seven data sets are taken.No hydraulic model is developed by the RJHAB and the model runs on a spreadsheet with a microcomputer.The IFIM models can generate estimates of equivalent optimum habitat called weighted usable areas (WUA's)with any flow within their calibration range,while the RJHABmodel only cal culates WUA I S at di scharges for whi ch measurements are taken.Therefore,it is necessary to interpolate between point measurements generated by the RJHAB model.The RJHAB model does have the advantage of being able to run in areas heavily influenced by mainstem backwater or sloughs with flows less than 5 cfs.The measure- ments and data analysis for the RJHAB model were taken by different investigators than those who took the IFIM measurements and analyzed them. The RJHAB model uses measurements at an additional upper transect within each of the sites.This upper area was very similar to lower sections of the site,and therefore would not change comparabil i ty of the two C-1 - methods.The IFIM presents resul ts of the ana lysi s on the basi s of a 1000 foot reach,while the RJHAB model presents WUA's for the site. Therefore,the length of each site as used in the RJHAB model was calculated and WUA's were adjusted to the basis of a 1000 foot reach. At Island Side Channel,two additional partial transects were put in for IFIM analysis of the site (see Appendix D),and no RJHAB measurements were taken at these transects;A trial run which minimized the effect of these two additional transects showed only very minor changes in WUA. RESULTS An IFG-2 IFIM model was run at Island Side Channel and hydraulic data were collected at a side channel flow of 338 cfs (Appendix D).At Trapper Creek Side Channel,hydraulic data for an IFG-4 IFIM model were collected at flows of 16,32,and 389 cfs.Habitat data for the RJHAB model were collected four times at Trapper Creek Side Channel and five - times at Island Side Channel and the RJHAB models at both sites were evaluated as "good"(Table 6).. The modelled response of area at the Trapper Creek and Island side channel sites to changes in discharge was almost identical for both the IFIM and RJHAB modelling techniques (Appendix Figure C-1).Differences in areas below the overtopping flow at Island Side Channel are probably due to the IFIM not being able to model flows below 5 cfs while the RJHAB WUA was measured at a flow of less than one cfs.Other differ- ences are readily attributable to sampling error.Since juvenile chinook and chum salmon are the two salmon species which make the heaviest use of side channels for rearing,only WUA results from these two species will be presented here. At Trapper Creek Si de Channel,the shape of the WUA curves for both speci es were basi cally the same for both mode"i ng methods (Appendix Figure C-2).The RJHAB model appears to consistently underestimate the amount of WUA in comparison to the IFIM model.The underestimation of WUA by the RJHAB model leads to smaller habitat indices although the shapes of the habitat index curves are similar for both techniques (Appendix Figure C-3). At Island Side Channel,on the other hand,WUAs from the two modelling methods do not compare closely (Appendix Figure C-4).The chinook and chum WUA response curves look more similar to each other than do the modelling techniques.Peaks in WUA for the RJHAB model occur at approx- imately 40,000 cfs while the IFIM model predicts a peak WUA at approxi- mately 60,000 cfs.The IFIM model does predict a chinook salmon WUA of 6,230 ft 2 to 6,600 ftz at side channel flows of 6 to 11 cfs which corresponds to the peak in the RJHAB model where a measurement was taken at a side channel flow of approximately 10 cfs. When habitat indices are calculated for both methods at Island Side Channel,differences between the two techniques appear smaller (Appendix Figure C-5).The RJHAB model shows a peak habitat index for chinook salmon at approximately 39,000 cfs which the IFIM model would also show at side channel flows of 6 to 11 cfs.Chum habitat indi ces for both C-2 ...., - Appendix Figure C-l.Comparison of site areas calculated with the RJHAB and IFIM modelling techni~ues for the Trapper Creek and Island Side Channel study sites. C-3 TRAPPER CREEK SIDE CHANNEL CHINOOK WEIGHTED USABl:..E AREA - 11 10 9 8 x loJ'fi'7 0."Zc;;:a 6~~m~5~.... 4- 3 2 10 c 1Fl... +RJHAB 30"50 70 .... CHUM WEIGHTED USABLE AREA 60 50 40 X Wfi"0."~a !<!l 30 !:::olDt:~-- 20 10 0 10 3d SO 70~OUSandS~" IIllAINSTE...DISC ARGE AT UNSHINE (cfs) Appendix Figure C-2.Comparison of weighted usable areas calculated with the RJHAB and IFIM modelling techniques for juvenile chinook and chum salmon at Trapper Creek Side Channel. 1984. C-4 - TRAPPER CREEK SIDE CHANNEL 0.12 CHINOOK HA81TAT INDICES ..-"" 0.11 0.1 a 1Fl.. 0.09 +RJHAB 0.08 xw 0.07Cl ~ !<0.06 I-' iii 0.05~ 0.04 0.03 0.02 0.01 0 10 50 70 -CHUM HA81TAT INDICES 0.7 <~0.6 a 1Ft.. +RJHAS 0.5-xw Cl 0.4~ !<~0.3 0.2 r 0.1 0 10 30 50 70 ~usondS~MAINSTEM DISC GE AT UNSHINE (cfs) -Appendix Figure C-3.Comparison of habitat indices calculated with the RJHAB and IFIM modelling techniques for juvenile chinook salmon at Trapper Creek Side Channel,1984. C-5 ISLAND SIDE CHANNEL CHINOOK WEIGHTED USABLE AREA 7 6 5 XW'ii'4o-g ~a 1-01~5 3aJf=~'"' 2 a IFlM +RJHAB - 7030 0;----,----..----,-----,---.,----,-----1 10 CHUM WEIGHTED USABLE AREA 42 40 38 36 34- l:k-32 0"z"O 30-6 ~~28!::~~c 26 24- 22 20 18 16 10 a IFlt.l +RJHAS 30 50 70 (Thousands) t.lAlNSTEt.I DISCHARGE AT SUNSHINE (cfs) Appendix Figure C-4.Comparison of weighted usable areas calculated with the RJHAB and IFIM modelling techniques for juvenile chinook and chum salmon at Island Side Channel,1984. C-6 ISLAND SIDE CHANNEL CHINOOK HASffAT INDICES 0.11 0.1 a FlM0.09 ~+RJHA8 o.oa )(w 0.070..-!;; !<0.06 ~0.05--0.04 0.03 .- 0.02 0.01 10 30 50 70 CHUM HA8lTATINDICES a lFlM +RJHA8 .- 30 50 70 (Thousonds) MAlNSTEM DISCHARGE AT SUNSHINE (cfs) Appendix Figure C-5.Comparison of habitat indices calculated with the RJHAB and IFIM modelling techniques for Juvenile chinook and·chum salmon at Island Side Channel,1984. C-7 techniques decrease after overtopping although the RJHAB habitat indices drop off more steeply. DISCUSSION The two modell"j ng methods compared very favorably at cal cul ati ng areas within the two sites.The shape of the chum and chinook WUA and habitat index responses at Trapper Creek Side Channel were very similar.The RJHAB model consistently underestimated WUA in comparison to the IFIM model.This is probably due to the RJHAB model not taking into account the area between-the shoreline cell and the cell located one-third of the way across the channel.This area was often marginal habitat with barely suitable velocities. At Island Side Channel,large differences in WUA can also be attributed, in part,to the RJHAB model not taking into account peripheral marginal habitat more than six feet from shore.This difference is also reflect- ed in the habitat indices where the proportion of usable area drops off more quickly for the RJHAB model.The differences in WUA below the overtopping flow can be attributed to the fact that the IFIM model does not run at flows less than five cfs while actual flows at discharges below the overtopping one are less than one cfs (Quane et al.1985). The effects of sampling errors in data collection on WUA estimates from both the RJHAB and IFIM techniques are unknown.Since many more meas- urements are taken for the IFIM,it should be less susceptible to sampling errors.Because only one IFIM measurement was taken at Island Side Channel at a flow of 338 cfs,however,the reliability of modelling flows as small as 5 cfs is unknown.It seems reasonable to assume that an IFG-4 model at Island Side Channel would have given somewhat differ- ent results than did the IFG-2 model.The RJHAB model works well in situations where the primary effect of discharge is due to backwater and the IFIM model cannot be used or works poorly. In summary,the RJHAB model generally gives lower WUA estimates than does the IFIM methodology.Also peaks in WUA are often narrower for the RJHAB model.Both models show the same general trends in the habi tat indices for chum and chinook salmon although the RJHAB model is more sensitive to increases in velocity and depth which decrease the habitat indices more quickly.Since the habitat indices for both sites cal- culated using both techniques are not appreciably different,analysis of trends and optimal flows by use of habitat indices would lead to similar cone 1us ions us i ng both methods.Compa ri sons of the I FIM wi th other instream flow methodologies have also shown differences in output,and no one method has yet been proven best (Annear and Conder 1984). C-8 - - - ,... LITERATURE CITED Annear,T.C.,and A.L.Conder.1984.Relative bias of several fish- eri es i nstream flow methods.North Ameri can Journal of Fi sheri es Management 4:531-539. Bovee,K.D.1982.A guide to stream habitat analysis using the in- stream flow incremental methodology.Instream Flow Information Paper.No.12.U.S.Fish and Wildlife Service.FWS/OBS-82/26. Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt. S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.;:-1984.Resi- dent and juveni le anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Prepared for the Alaska Power Authority. Anchorage,Alaska. Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L. Crawford,and P.M.Suchanek (eds:T:1984.Resident and juvenile anadromous fish investigations (May -October 1983).Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.Report No.2.Prepared for the Alaska Power Authority.Anchorage, Alaska. Milhous,R.T.,D.L.Wegner,and T.Waddle.1981.User l s guide to the physical habitat simul ation system (PHABSIM).Instream Flow Information Paper 11.U.S.Fish and Wildlife Service FWS/OBS-81/43. Quane,T.,P.Morrow,1.Queral,T.Keklak,and T.Withrow.1985. Technical memorandum in support of Task 14 (Lower River Resident and Juvenile Anadromous Fish Studies)Alaska Department of Fish and Game Susitna Aquatic Studies.Anchorage,Alaska. C-9 - - I- ,, APPENDIX D HYDRAULIC MODELS FOR USE IN ASSESSING THE REARING HABITAT OF JUVENILE SALMON IN SIX SIDE CHANNELS OF THE LOWER SUSITNA RIVER ..- .- -- ,...., APPENDIX D HYDRAULIC MODELS FOR USE IN ASSESSING THE REARING HABITAT OF JUVENILE SALMON IN SIX SIDE CHANNELS OF THE LOWER SUSITNA RIVER By: James Anderson, Andrew Hoffmann,and Jeffrey Bigler of Alaska Department of Fish and Game Susitna River Aquati c Stud;es Program Third Floor,Michael Building 620 East Tenth Avenue Anchorage,Alaska 99501 ABSTRACT Six side channels (Island,Mainstem West Bank,Circular,Sauna,Sunset, and Trapper Creek)in the lower reach of the Susitna River were evalu- ated using an Instream Flow Incremental Methodology (IFIM)physical habitat simul ati on (PHABSIM)mode,l1 i ng approach to describe the effects that site flow and mainstem discharge have on rearing juvenile salmon habitat.These sites were thought to contain potential habitat for rearing juvenile salmon and were chosen to range greatly in size,shape, and overtopping discharge. Si x hydraul ic simul ati on model s (either IFG-2 or IFG-4)were calibrated to simulate depths and velocities associated with a range of site- specific flows at the six modelling study sites.Comparisons b~tween correspoDding sites of simulated and measured depths and velocities indicated that the models provide reliable estimates of depths and velocities within their recommended calibration ranges. The recommended of ranges of mainstem Susitna River discharge over which these models can hydraulically simulate the habitat of rearing juvenile salmon are:Island Side Channel from 35,000 to 70,000 cfs mainstem discharge;Mainstem West Bank Side Channel from 18,000 to 48,000 cfs; Circular Side Channel from 36,000 to 63,000 cfs;Sauna Side Channel from 44,000 to 63,000 cfs;Sunset Side Channel from 32,000 to 67,000 cfs;and Trapper Creek Side Channel from 20,000 to 66,000 cfs. D-i ..... TABLE OF CONTENTS ABSTRACT ••••••.•.• TABLE OF CONTENTS ..••.•••.• LIST OF APPENDIX FIGURES ••. LIST OF APPENDIX TABLES •. D-i D-ii D-iv .D-viii INTRODUCTION •..................................................... METHODS •••.•••• 0-1 D-1 Approach •••.•-Analytical ..........................,..D-1 Study Site Selection. ......................--General Techniques for Data Collection. D-3 D-4 Genera 1 Techniques for Calibration •.D-7 Island Side Channel ••• - General Techniques for Verification •••••••••••..•.•••••.•..•• R-ESULT'S ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• ....................................... Site Description .......•................... Calibration •••••• Verification •••.. Application ...•.•••.•.....o ••••••• D-10 0-10 D-13 D-13 D-19 D-19 0-25 Mainstem West Bank Side Channel ..............'.D-26 Site Description •••• Calibration ..•. Verification .•• Application ••••••.• Circular Side Channel •. Site Description •. Calibration ••••••• Verification .••... Application ••••••• ...................... .................. D--;i 0-26 D-32 D-35 D--35 D-39 0-39 0-39 D-46 D-49 TABLE OF CONTENTS (Continued) Sauna Side Channel ••••••••• Site Description ... Calibration •••••••• Verification •••.•.•.••..• Application •••• 0-51 0-51 0-51 0-60 0-62 Sunset Side Channel .••••••••••••••••••II •••••••••e •••••••••••tJo 0-64 Site Description ..••••••..•. Calibration .••••••••• Verification ••.•. Application •.•...••••• Trapper Creek Side Channel Site Description .••.••••• Calibration •••••.•••••••• Verification •••••• Application •.•.•.. 0-64 0-70 0-74 0-74 0-77 0-77 0-77 0-83 0-87 SUMMARy •.•••.•••.• ACKNOWLEDGEMENTS. •••••••••••O ••••••c."'•••••••••O.CiI ••••0-87 0-88 LITE RATURE CITED •.•••••s ••~• '••III ••Co ••••••••••••••••e e ••CI •ill _.'••••••• D-iii 0-89 ...., LIST OF APPENDIX FIGURES Appendix Figure Title - D-1 Location of the six IFG hydraulic modelling sites in the lower Susitna River......................0-5 D-2 Overview of Island Side Channel (RM 63.2)••••.•••••••.0-14 0-3 Location of Island Side Channel study site (RM 63.2}•.•...•....•.•.....,................•......•...0-15 0-4 Comparison of rating curves for Island Side Channel transect 6 (Q site)(from Quane et a 1.1985),-.. . . . . . . . . . . . . . . . . . . . . . . . . . .D-16 D-5 Cross section of transects 1,lA,2,and 3 at Island Side Channel (adapted from Quane eta.'.1985)'..................0-17 "'"' 0-6 Cross section of transects 4,4A,5,and 6 at Isl and Si de Channel (adapted from Quane et al.1985)-- '0-18 -- 0-7 0-8 Compari son of observed and predi cted water surface profiles from cal ibrated model and surveyed thalweg profile at Island Side Channel (adapted from Quane et al.1985)••••••••••••••D-20 Application range of the calibrated hydraulic model at Island Side Channel..........................0-22 (RM 74.4)•..••....•••.•.••.••••••••.•••~.•...••.•.'•. •.D-27 Location of Mainstem West Bank Side Channel stu dy site (RM 74.4)••••••••••••••••••••••••••••. •. ••• D-28 Comparison of rating curves for Mainstem West Bank Side Channel transect 1 (Q site)(from Quane et ale 1985).......•............................D-29 D-9 Comparison of observed and predicted.veloc- ities from the IFG-2 hydraulic model at Island Side Channel,using two flows at the transect 1 discharge site ••••••••••••••••••••••••.••.•D-23 Comparison of observed and predicted veloc- ities from the IFG-2 hydraulic model at Island Side Channel,using two flows at the transect 6 discharge site •••••••~•••••••••••••••••••••D-24 Overview of Mainstem West Bank Side Channel - 0-10 P"'" D-11 0-12 ~ 0-13 I""" D-iv LIST OF APPENDIX FIGURES Appendix Figure Title 0-14 Cross section of transects 1,2,and 3 at Mainstem West Bank Side Channel (adapted from Quane et ale 1985).................................•..0-30 - Cross section of transects 3A and 4 at Mainstem West Bank Side Channel (adapted from Quane et ale (1985)••••••.••••~••.••.••••ee.e~e •••e •••0-31 Comparison of observed and predicted water surface profiles from cal ibrated model and surveyed thalweg at Mainstem West Bank Side Channel (adapted from Quane et al.1985).•.••..••.••.•0-33 0-17 Application range of the calibrated hydraulic model at Mainstem West Bank Side Channel ......•••.••.•0-36 0-16 0-15 0-18 Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraul ic model at Mainstem West Bank Side Channel •.••.•......•.••..••••••••••.•..•.•••.••.•••...0-37 """ 0-19 Overview of Circular Side Channel (RM 75.3)•.......••.0-40 0-20 Comparison of ratin9 curves for Circular Side Channel transect 4 {from Quane et al.1985)..•.••...•.0-41 0-21 Location of Circular Side Channel study site (RM 75.3)..e _.........................................0-42 0-22 Cross section of transects 1,2,and 2A at Circular Side Channel (adapted from Quane et al.1985)"..e1c •••••,.,••e •••••t1 ••••••••••D-43 ·0-23 Cross section of transects 3,4,and 5 at Circular Side Channel ..•••.••••..••..•.••...•.••••.•.•·0-44 0-25 0-24 Comparison of observed and predicted water surface profi 1es from cali brated model and surveyed thalweg profile at Circular Side Channel (adapted from Quane et al.1985}..••••.•..•.••0-45 Application range of the calibrated hydraulic model at Circular Side Channel .•......•..••...•..•....D-48 - 0-26 Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraulic model at Circular Side Channel •.•••.•.•••..•0-50 0-27 Overview of Sauna Side Channel {RM 79.8).•.•.•...••...0-52 D-v - LIST OF APPENDIX FIGURES Appendix Figure Title - 0-28 Comparison of rating curves from Sauna Side Channel transect 2 (from Quane et ale 1985).......•..•0-53 0-29 Locations of Sauna Side Channel study site (RM 79.8)-..•..•.......•...••.....•........0-54 0-30 Cross.section of transects 1~2~3~and 4 at Sauna Side Channel (adapted from Quane et al. 1985).... ................... . . . . . .................... ..0-55 0-31 Comparison of observed and predicted water surface profi 1es from calibrated model and surveyed thalweg profile at Sauna Side Channel (adapted from Quane et al.1985)••..••••••••.•0-56 0-32 Application range of the calibrated hydraulic mode 1 at Sauna Si de Channe 1.. . . . •. . . . . . •. . . . . . •••. . . . .0-61 0-33 Comparison of observed and predicted veloc- ities from the IFG-2 hydraulic model at Sauna Side Channel using two flows at the discharge transect..••••.••. . ..••••••.•••.•. ..•.•.•••. •••••. . . . •0-63 D-34 Overview of Sunset Side Channel (RM 86.4)••••••••..•••D-65 0-35 Comparison of rating curves from Sunset Side Channel at transect 1 (from Quane et al. 1985)~'0-66 0-36 Location of Sunset Side Channel study site (·RM 86.9)•.•••••••••••••••.•._••••.••••••~.•••••.•••••_.0-67 D-37 Cross section of transects 0,1~2,and 3 at Sunset Si de Channel (adapted from Quane et a 1.1985).............................................D-68 0-38 Cross section of transects 4~5~and 6 at Sunset Si de Channel (adapted from Quane et al.1985)0-69 D-39 Compari son of observed and predicted water surface profiles from calibrated model and surveyed thalweg profile at Sunset Side Channel (adapted from Quane et al.1985)••.•••.•••••••0-71 0-40 Application range of calibrated hydraulic model at Sunset Side Channel ••••••••.••.•••••••.••••••0-75 D-vi LIST OF APPENDIX FIGURES (Continued) Appendix Figure Title 0-41 0-42 0-43 0-44 0-45 0-46 0-47 0-48 Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraulic model at Sunset Side Channel .....••••••.•••.0-76 Overvi ew of Trapper Creek Si de Channel (RM 91.6)0-78 Comparison of rating curves from Trapper Creek Side Channel transect 4 (from Quane et al.1985).,...,0-79 Location of Trapper Creek Side Channel study site (RM 91.6).....G ••e •••••••e •••••••••••••••o ••••e.o 0-80 Cross section of transects 1~2~3,and 4 at Trapper Creek Side Channel (adapted from Quane et al.1985)........•.•.......•..........•.•....0-81 Comparison of observed and predicted water surface profiles from calibrated model and surveyed thalweg profi 1e for Trapper Creek Side Channel (adapted from Quane et ale 1985).Cl •.•••••••••••.•••••••..•..••I1 0-82 Application range of the calibrated hydraulic model at Trapper Creek Side Channel ..........•....•.•.0-85 Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 model at Trapper Creek Side Channel ...••....•.......•.0-86 D-vii .~, - ~ ) - - - LIST OF APPENDIX TABLES Appendix Table Title 0-1 The six lower river IFG modelling sites with corresponding river mile location.....................0-6 D-2 Percent cover and cover type categories...............D-8 0-3 0-4 Substrate classifications ......••.••••....o.0 •••••••0.0-9 The six lower river side channel IFG model- ling sites with type of hydraulic model used, dates calibrations flows measured,and corresponding site specific flows and mainstem discharges for the open-water period in 1984.0 0 0 •••0 •~0 •••0 • 0 ••••••0 • 0 • 0 • 0 ••o ••00 0 0 •••0.• • •0-12 - 0-5 Comparison of field measured and model predicted water surface elevations at the calibration flow of 338 cfs for Island Side Channel ..o.••0 •••0 • 0 0 00 0 0 • 0 0 •••0 • 0 • 0 • 0 • 0 0 0 0 00 • 0 ••0 ••0 •0-21 0-6 Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Mainstem West Bank Side Channel hydraulic model.0 0 •••0 ••0 •••••••0 0 ••o'0 0 o.0 o.0 0-34 0-7 The statistical results used to evaluate the predictive ability of the four lower river IFG-4 hydraul ic model so 0 0 00 .....0 ....0 0 ...0"0 0"0....0-38 0-8 Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Circular Side Channel hydrauli c model 0 •••0.0 • 0 0 • 0 • 0 • 0 0 ••0 0 •••0 0 •••0 ••• • • • • • •0-47 0-9 Comparison of fiel d measured and model predicted water surface el evations at the calibration flow of 52 cfs for Sauna Side Channel .••••••..•.0"0 ••••••••••••••••••••••••o.••.•••0-58 0-10 The effects of the backwater at Sauna Side Channel,information obtained from transect2.0 0...............0-59 0-11 Comparison between observed and predicted water surface elevations,discharges and velocities for 1984 Sunset Side Channel hydraul ic model ••0 ••••••••••o..•....•...••..•.... •..•.0-72 0-12 Differences between stages of zero flow input into the model and Quane et ale (1985) thalweg survey at Sunset Side Channel •..•...•........•0-73 D-viii LIST OF APPENDIX TABLES (Continued) Appendix Table Title 0-13 0-14 Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Trapper Creek Side Channel hydraulic model •.........•••....••.......•.••.0-84 Summarization of the range of mainstem discharges that the hydraulic models can simulate for the rearing habitats of salmon at the six lower river IFG modelling sites •....••.....0-87 D-ix ,~ -- - !~ I"""' i INTRODUCTION About 40%of the annual discharge of the lower Susitna River at the Parks Highway bridge originates from the mainstem Susitna River above the confluence of the Talkeetna and Chulitna Rivers (Acres 1982).Thus, operation of the proposed hydroelectric project will alter the natural flow regime of this lower river reach beyond the normal variations in flow which occur naturally during the open-water season. One of the predominant aquatic habitat types in this lower river reach which may be affected by such flow alterations are side channels.Side channel areas in this river reach currently provide habitat for rearing juvenile salmon.The quantity and quality of juvenile salmonid rearing habitat in side channels in this river reach is dependent on a multitude of interrelated habitat variables,including water depth and velocity, which are intimately related to mainstem discharge. This appendix presents results of the physical habitat modelling simu- lation efforts that Alaska Department of Fish and Game (ADF&G)Su Hydro personnel conducted during the open-water season of 1984.The objective of the study was to provide calibrated hydraulic simulation models for selected lower river juvenile salmon habitat modelling study sites.The approach of the study was to apply a methodology which used water depth and velocity as the dominant hydraulic variables to quantify the responses of rearing habitat to changes in site flow and mainstem discharge.The methodology used was the system developed by the U.S. Fish and Wildlife Service (USF&WS)Instream Flow Group (lFG)called the Instream Flow Incremental Methodology (IFIM)Physical Habitat Simulation (PHABSIM)modelling system (IFG 1980,Bovee 1982).The calibrated hydraulic simulation models will be utilized to assess how site flows and mainstem discharge affect juvenile salmon rearing habitat in side channels of the lower Susitna River. METHODS Analytical Approach A common methodology used for assessing habitat responses to flow variations is the IFIM,·PHABSIM modell ing system.The IFIM,PHABSIM modelling system is a collection of computer programs used to simulate both the available hydraulic conditions and usable habitat at a study site for a particular species/life phase as a function of flow.It is based on the theory that changes in riverine habitat conditions can be estimated from a sufficient hydraulic and biological field data base. It is intended for use in those situations where flow regime and channel structure are the major factors influencing"river habitat conditions. The modelling system is based on a three step approach.The first step uses field data to calibrate hydraulic simulation models to forecast anti ci pated changes in physi ca1 habi tat va ri ab 1es important for the species/life phase under study as a function of flow.The second step involves the collection and analysis of biological data to determine the behavioral responses of a particular species/life phase to important physical habitat variables.This information is used to develop 0-1 weighted behavioral response criteria curves (e.g.,utilization curves, preference curves,or suitability curves).The third step combines information gained in the first two steps to calculate weighted usable area (WUA)indices of habitat usability as a function of flow for the species/life phase under study. Hydraulic modelling is of central importance to the PHABSIM system.The primary purpose of incorporating hydraulic modelling into the analytical approach is to make the most efficient use of limited field observations to forecast hydraulic attributes of riverine habitat (depths and veloc- ities)under a broad range of unobserved streamflow conditions. The IFG developed two hydraulic models (IFG-2 and IFG-4)during the late 1970's to assist fisheries biologists in making quantitative evaluations of effects of streamflow alterations on fish habitat.The IFG-2 hy- draulic model is a water surface profile program that is based on open channel flow theory and formulae.The IFG-2 model can be used to predict the horizontal distribution of depths and mean column velocities at 100 points along a cross section for a range of streamflows with only one set of field data.The IFG-4 model provides the same type of hydraulic predictions as the IFG-2 model,but it is more strongly based on field observations and empiricism than hydraulic theory and formulae. Although a minimum of two data sets are required for calibrating the IFG-4 model,three are recommended.Either model can be used to fore- cast depths and vel oeiti es occurri ng ina stream channel over a broad range of streamflow conditions. The IFG-4 model,which is based upon a greater number of observed sets of field data (i.e.flow levels),generally can be used to model a greater range of flow conditions than the IFG-2 model.Additionally, since the IFG-4 model is more dependent upon observed depths and veloc- ities than the IFG-2 model,predicted depths and velocities can be directly compared with the observed values.This comparison is a useful tool for verifying the models. Both models are most applicable to streams of moderate size and are based on the assu~ption that steady flow conditions exist within a rigid stream channel.A stream channel is rigid if it meets the following two criteria:(1)it must not change shape during the period of time over which the cal ibration data are collected,and (2)it must not change shape while conveying streamflows with-in the range of those that are to be simulated.Thus a channel may be "r igid"by the above definition, even though it periodically (perhaps seasonally)changes course. Streamflow is defined as "s teady ll if the depth of flow at a given location in the channel remains constant during the time interval under consideration (Trihey 1980). In this analysis,all streamflow rates were referenced to the average daily discharge of the Susitna River at the U.S.Geological Survey (USGS)stream gage at Sunshine,Alaska (station number 15292780).This location was selected as the index station primarily because it is the gage located near the center of the river segment that is of greatest interest in this particular analysis.The target mainstem discharge range for data collection was from 12,000 to 75,000 cfs. 0-2 - - - ...." .-, i """ - Site specific streamflow data collected during 1984 provided the basis for correlating flow through the various study sites to the average daily streamflow of the Susitna River at the Sunshine gage.Detailed site specific channel geometry and hydraulic measurements provided the necessary data base to calibrate hydraulic models for each study site. Informati on for two .other physical habitat vari ab 1es,substrate and cover,were also coll ected.Substrate was not incorporated into the models at this time,but cover,an important variable in assessing the habitat quality for most rearing salmon juveniles,was • These data and hydraul ic model s make up the physical habitat component of the PHABSIM analysis.For a given discharge of the Susitna River at Sunshi ne,the flow through each study site can be determi ned and site specific hydraulic conditions (velocity and depth)can be predicted. The results based on velocity,depth,and cover may be used to forecast the effects of mainstem discharge on the weighted usable area for juvenile rearing salmonids of these modelled side channel habitats. Study Site Selection Two basic approaches are comnonly used for selecting study sites to be evaluated using the IFIM PHABSIM modelling system:the critical and representative concepts (Bovee and Milhous 1978;Trihey 1979;Bovee 1982).Application of the critical concept requires knowledge of a stream's hydrology,water chemistry,and channel geometry in addition to rather extensive knowledge of fish distribution,relative abundance,and species-specific life history requirements.Criteria for application of the representative concept are less restrictive,enabl ing this concept to be used when only limited biological information is available or when critical habitat conditions cannot be identified with any degree of certainty. In the critical concept,a study area is selected because one or more of the physical or chemical attributes of the habitat are known to be of critical importance to the fish resource.That is,recognizable phys- ical or chemical characteristics of the watershed hydrology,instream hydraulics,or water quality are known to control species distribution or relative abundance within the study area.Because of this,an eval- uation of critical areas will provide a meaningful index of species response in the overall critical study area. The representative reach concept acknowledges the importance of physical habitat variables throughout the entire study stream for sustaining fish populations.Thus,under this approach,study areas are selected for the purpose of quantifying relationships between streamflow and physical habitat conditions important for the species/life phase under study at selected locations (representative reaches)that collectively exemplify the general habitat characteristics of the entire river segment. For thi s study,an adaptation of the representative concept was the approach used to assess how mainstem discharges affect the rearing habitat of juvenile salmon in side channel complexes.The six lower river IFG study sites are most representative,morphologically,of D-3 intermediate side channels and of the habitat type designation,sec- ondary side channel as described by Ashton and Klinger-Kingsley (1985). The results from these six IFIM-PHABSIM models are probably most appli- cable to these types of areas in segments I and II of the lower Susitna River.This segmentation of the lower river is also described in Ashton and Klinger-Kingsley (1985).The six study sites were chosen by ADF&G .Su Hydro Resident and Juvenile Anadromous (RJ)project personnel in conjunction with ADF&G Su Hydro Aquatic Habitat and Instream Flow Study (AH)project and E.Woody Trihey and Associates (EWT&A)personnel from lower river side channels which met the following basic criteria: 1.The sites were chosen to range greatly in size,shape,and overtopping discharge; 2.The sites were thought to contain potential habitat conditions for rearing juvenile salmon; 3.The sites were judged by AH project and EWT&A personnel to be readily modelled using the IFIM methodology; 4.The sites were accessible by boat at normal mainstem dis- charges during the open-water season;and, 5.The sites were above Kashwitna landing and therefore much easier to sample for logistical purposes. The six sites chosen for modelling complemented other sites modelled using another habitat model (see main text).All of the six sites were side channels,the majority of potential habitat in the lower river is composed of this habitat.Much of the other habitat ;s difficult to model with the IFIM methodology because it is affected primarily by mainstem backwater.Appendix Figure 0-1 shows the location of each of the six sites selected for study,the corresponding river mile location is presented in Appendix Table D-l. General Techniques for Data Collection A study reach was selected for detailed evaluation in each of the six side channel sites.The length of the reach was determined by placing enough transects within the area to adequately represent the major macrohabitat types of the particular side channel area. Transects were located within each study reach following field methods described in Bovee and Milhous (1978)and Trihey and Wegner (1981),and were located to facilitate collection of hydraulic and channel geometry measurements of importance in evaluating flow effects on salmon rearing habitat.Field data were obtained to describe a representative spectrum of water depth and velocity patterns,cover,and substrate composition at each side channel reach. The number of transects estab 1i shed at the study reaches vari ed from four to eight.The end points of each transect were marked with 3D-inch steel rods (headpins)driven approximately 28 inches into the ground. The elevation of each headpin was determined by differential 0-4 ..... i i 10 I PARKS HIGHWAY BRIDGE:~~::::::::L::'USGS STATION (15292780) rRAPP£R CREEK SIO£CHANNEL COOK INLET - - .- i Appendix Figure 0-1.Location of the six IFG hydraulic modelling sites in the lower Susitna River. 0-5 Appendix Table 0-1.The six lower river IFG modelling sites with corresponding river mile location. Side Channel Site River Mil e ~ Island Side Channel 63.2 ......, Mainstem West Bank Side Channel 74.4 Circular Side Channel 75.3 Sauna Side Channel 79.8 ~ Sunset Side Channel 86.9 Trapper Creek Side Channel 91.6 - .... ~I - D-6 I~ r leveling using temporary benchmarks set at assumed elevations of 100.00 feet. Cross section profiles at each transect were measured with a 1evel , survey rod,and fi bergl ass tape.Hori zonta 1 di stances were recorded to the nearest 1.0 foot and streambed elevations to the nearest 0.1 foot. Water surface elevations at each cross section in the study site were determined to the nearest 0.01 foot by differential leveling or by reading staff gages located on the cross section. Streambed elevations used in the hydraulic models were determined by making a comparison between the surveyed cross section profile and the cross section profiles derived by subtracting the flow depth measure- ments at each cross section from the surveyed water surface elevation at each calibration flow (Trihey 1980). A longitudinal streambed profile (thalweg profile)was surveyed and plotted to scale for each modelling site (Quane et al.1985). The water surface elevation at which no flow occurs (stage of zero flow) at each cross section in the study site was determined from the stream- bed profile.If the cross section was not located on a hydraulic control,then the stage of zero flow was assumed equal to that of the control immediately downstream of the cross section. Discharge measurements were made using a Marsh-McBirney or Price AA velocity meter,topsetting wading rod,and fiberglass tape.Discharge measurements were made using standard field techniques (Buchanan and Somers 1969;Bovee and Milhous 1978;Trihey and Wegner 1981).Depth and velocity measurements at each calibration flow were recorded for the same respective points along the cross sections by referencing all horizontal measurements to the left bank headpin. Cover and substrate values were also determined for each cell along model 1ing transects..Methods described in Suchanek et al.(1985)were used to code cover (Appendix Table 0-2).Substrate categories were clas- sified by visual observation employing the substrate classifications presented in Appendix Table 0-3.The distribution of various substrate types was indicated on field maps.Substrates were classified using a single or dual code.In those instances that a dual code was used,the first code references the most predominant (i.e.,70%rubble/30%cobble =9/11). General Techniques for Calibration The calibration procedure for each of the hydraulic models was preceded by field data collection,data reduction,and refining the input data. The field data collection entailed establishing cross sections along which hydraulic data (water surface elevations,depths,and velocities) were obtained at each of the different calibration flows.The data reduction entailed determining the streambed and water surface ele- vations,velocity distribution,the stage of zero flow for each cross section,and determining a mean discharge for all the cross sections in the study site.A model was considered calibrated when:1)the D-7 Appendix Table D-2.Percent cover and cover type categories. .., I Cover Type Code %Cover Code silt,sand (no cover)1 0-5 .1 emergent vegetation 2 6-25 .2 aquatic vegetation 3 26-50 .3 """'i 1-3"gravel 4 51-75 .4 -3-5"rubble 5 76-100 .5 5 11 cobble,boulder 6 - debris 7 overhanging riparian vegetation 8 undercut bank 9 - 0-8 - Appendix Table D-3.Substrate classifications. Substrate Type Part;cle Size Classification Sil t Silt 1 2 Sand Sand 3 4 Small Gravel 1/8-1"5 6 Large Gravel 1_31/7 8 Rubble 3-5"9 10 ~, Cobble 5-10"11 12 Boulder 10",13 ~ - 0-9 majority of predicted water surface profiles were within ±0.05 ft of the observed elevations and 2)the majority of predicted velocities were within ±O.10 ft/sec of the measured velocities.A calibrated IFG-4 model gives velocity adjustment factors in the range of 0.9 to 1.1,and relatively few velocity prediction errors.The velocity adjustment factor is the ratio of the computed (observed)discharge to the predict- ed discharge. An IFG-2 model does not have velocity adjustment factors and is reviewed with the observed data before it is considered calibrated. For a more detailed explanation of the general techniques used for calibrating the IFG-2 and IFG-4 models in the lower river see Hilliard et a1.(1985). General Techniques for Verification The verification of how well each of these six hydraulic models simulat- ed their respective site flows was performed by the hydraulic engineers at EWT&A.The approach used to assess the qual ity of each model was based on two levels of criteria.The first was a qualitative evaluation of four separate sub-criteria.These sub-criteria were: 1.How well does the model conform to the IFG (Main 1978 and Milhous et ale 1984)and EWT&A (Hilliard 1985)guidelines? 2.How well does the extrapolation range of the model conform to the desired range? 3.Are the model s appropriate for the species and 1He stage being .considered? 4.How well do the ranges of depth and velocities of the fore- casted data conform to the ranges of depth and velocity of the suitability criteria curves being considered based on a ~visual~evaluation? After the first level of qualitative evaluation was performed,an overall rating was given to the various segments of each model.The ratings given were excellent,good,acceptable,and unacceptable. Figures depicting these rating are presented for each site in the results section.The second level of the verification process required a statistical analysis to evaluate the models calibration.It was only performed when the forecast capabilities of either the IFG-2 and IFG-4 model were not given an excellent rating in the level one evaluation. For a detailed explanation of the verification analysis see Hilliard (1985). RESULTS The results of the physical habitat simulation modelling studies are presented below by study site.The six lower river side channel IFG modelling sites with type of hydraulic model used,dates calibration flows were measured,and corresponding site specific flows and mainstem D-I0 ...... - ~, - - ,.... - .- discharges for the open-water period in 1984 are presented in Appendix Table 0-4.The following items are presented for each study site:(1) a general site description,(2)a summary of data collected,(3)a description of procedures used to calibrate the model,(4)the verifi- cation of the model,and (5)the recommended application of the model for each study site. 0-11 Appendix Table 0-4.The six lower river side channel IFC modelling sites with type of hydraulic model used,dates calibration flows measured,and corre- sponding site specific flows and mainstem discharges for the open water period in 1984. Mainstem Date Site Discharge Type of Calibration Specific at aSideChannelHydraulicFlowFlowSunshine Site (RM)Model Measured (cfs)(cfs)""'" Island Side Channel (63.2)IFC-2 July 25 338 56,100 Mainstem West Bank (74.4)IFG-4 September 2 450 32,000 September 20 310 30,500 September 25 6 19,600 Circular Side Channel (75.3)IFC-4 July 24 204 55,200 August 17 50 42,500 Sauna Side Channel (79.8)IFC-2 July 23 52 52,000 -Sunset Side Channel (86.9)IFG-4 July 22 496 57,800 August 17 127 42,500 Trapper Creek Side Channel (91.6)IFC-4 September 18 16 20,900 p:}, August 16 32 44,000 July 21 389 57,700 a Mainstem discharge determined from provisional USCS streamflow data from the stream gage ~ at Sunshine,Alaska (station number 15292780). - 0-12 - - - Island Side Channel (RM 63.2) Site Description Island Side Channel is located on the east bank of the main channel of the Susitna River at river mile (RM)63.2 (Appendix Figure D-2).This side channel is located downstream of a braided,vegetated floodplain and is not directly connected to the main channel Susitna River.It is approximately 0.7 miles in length with both the mouth and head portions adjoining side channel networks.Breaching flows in this side channel result from overtopping of the head by an adjoining larger side channel. Prior to breaching,flow in the side channel is small with a series of pools remaining (Quane et ale 1985). The IFG modelling site at Island Side Channel was 735 feet long and located in the lower portion of the side channel (Appendix Figure D-3). The site generally consists of a pool-riffle-pool sequence.Based on assessments by Quane et ale (1985),an area of backwater extends through the study site to a point at least 1,100 feet upstream from the mouth of the side channel at a nOQ.-breaching mainstem discharge of 35,000 cfs. During mainstem discharges of 38,000 to 66,700 cfs,the area of back- water extends throughout the study site. The right bank of the study site is about five feet high,and the bank is steep due to the effects of erosion.The primary riparian vegetation along this.bank ;s alder.There are two side pocket areas along this bank,which become slack water areas during higher site flows (400 cfs).In contrast,the left bank of the study site is a gently sloping depositional bank.The riparian vegetation on this bank is sparse consi sting primarily of shrub wi 11 ow. Substrate at the study site consists primarily of gravels and rubbles, with substrate changing to sand and silt in slackwater areas.The thalweg gradient of the side channel is 15.6 ft/mile (Quane et ale 1985).From an evaluation of field observations,aerial photography, and the stage/discharge relationship developed for this side channel,an initial breaching has been estimated to occur at a discharge of 34,000 cfs (Quane et al.1985). Based on a .review of available rating curves (Appendix Figure D-4)it was determined that the hydraulics within this side channel are directly controlled by mainstem discharges exceeding 35,000 cfs (Quane et ale 1985).A side channel streamflow of 43.5 cfs has been estimated to occur at a mainstem discharge of 35,000 cfs (Quane et al.1985). Eight cross sections were surveyed within this site during 1984 to define channel geometry (Appendix Figures D-5 &D-6).The upper two transects (5 and 6)were primarily located in pool habitat.Transects 4A and 4 primarily represent riffle habitat in the main portion of the channel.Transect 4A was placed as a partial transect originating from the right bank.It represents the 1arger of the two sl ack water areas in this reach.The four downstream most transects are primarily in pool type habitat.Transect 1A was also a partial transect,representing the smaller slack water area along the right bank. D-13 Appendix Figure 0-2. e River Mile 01-(__-===2....1°PO t'f· Overview of Island Side Channel (RM 63.2). 0-14 - ..... ..... - - .~-1 -1 ._--,1 -1 1 }1 ,1 -,1 CI I f-I U1 Appendix Figure 0-3.Location of Island Side Channel study site (RM 63.2). - ! - - - - t 1 ICIl... Ic;i l.:os I~ I~ WSEL Et89 feet) Q=43.5 ct. CiK •10-0·1431 IIlSlL •II)C.U3t pI iii .ft." ISLAND SIDE CHANN,EL TR6 GAGE 63,256 i'4------~-..L-...._-.__-............__ I 3=o ..J 1.L. 0' Wa::::> (J) <C W ~ 3S.aoo ~II !".JOG crl IlSIL •MoO.'",1.DIlI •II r'l._.'!II o.c •10·n .tI"0..I.UJt v'.too'" IIOt CGOfIlOUU 17'-S,S 15._~ft. WKL .Io-a·om ".lIot •II ,.2.0." ..J W (J) 3= ! -...a _ MAINSTEM DISCHARGE.SUNSHINE (x 1000 cfs), t•~I 81q,1on ':?I ';°1 ~ en I !WSEl=91.79 I 3=o ..J Ll..Q=68.8owa::::> (J) <C W :::E ,..,. 0'1 '0.... 'e __ MAINSTEM DISCHARGE,SUNSHINE (x 1000 cfs) Appendix Figure 0-4.Comparison of rating curves for Island Side Channel transect 6 (Q site)(from Quane et.al. 1985). 0-16 1.-)J ,1 1 )1 1 1 }.1 1 2802402001110'208040 CROSS SECTION IA STATION 0+44 lOll 104 103..102 CD 101CD-100 z:1111 0 lie i=117 <t 116 >.lIl1l1J -I il4 lJJ 93 l1J 112 >In...110 <t 1111-I IIJ 1111 I!:117 IIIl 1I11 02802402001110UO110 ............--=I "6 eft 40 CROSS SECTION I STATION Oi-OO lOll 104 '03..102II>10'...... '00 z:9D 0 DB ~97 DIl l1J 911 -I 114 l1J 113 W 92>III~110 -I IlII l1J IlllI!:117 IlS Illl 0 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet) DISTANCE FROM LEFT BANK HEADPIN (feet) '---..........<J 338 eft CROSS SECTION 3 STATION 2+55 2802402001110UO80·40 Cl I lOll lOllt--'104 CROSS SECTION 2 104......STATION 1+12103 ~'03;:102 ..102..II> CD 101 III 10'---100 -100 z:89 z:1111 Q 1111 0 1111...117 j:97 ~116 ~91l I&J IIl1 IIJ 1I11 -I 114 -I il4IIJ---558 e I&J113 113 lJJ 112 IIJ 92 >9'>III j:110 i=80<t<t 1111 -I BII-I I&JlJJ1111I!:1111 I!:87 87 liB Illl 811 115 0 40 110 120 1110 200 240 2BO 0 DISTANCE FROM LEFT BANK HEADPIN «feet) Appendix Figure 0-5.Cross section of transects 1,lA,2,and 3 at Island Side Channel (adapted from Quane et al.1985). 2110240200111012011040 '.....1 ..........l 338 ell CROSS SECTION 4A STATION 4+31 105 104 103 Z 102 tlJ 101tlJ-100- Z 1111 2 1111 ~117«911>115W ..J 114 W 113 W 112 >111 ~110«1111..Jw 1111 0::117 1111 115 0280240200111012011040 '---00::I sse.to CROSS SECTION 4 STATION 3"62 10:1 104 103 Qi 102 tlJ 101--100 Z 1111 0 1111 5 117 >1I11 W 115 ..J 114W 113 W 112>IIIi=«110 ..J 1111 W 11110::117 1I11 115 0 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet I DISTANCE FROM LEFT BANK HEADPIN (fellt I 2110240200111012080 .....",sse .11 40 CROSS SECTION 6 STATION 7+35 0 I 105 10:1......104 CROSS SECTION 5 10400 103 STATION 5+65 103--..102 -;102•101 101••-100 .-100~ Z 1111 1111 ZQ11110 118 ~87 i=117 «1I11 «1111>115 >115WW..J 114 sse .11 ..J 114 W 113 W 113 W 112 W 112 >111 >111 i=80 j:110«119 «119..J ..JW1111W 1111 0::117 0::117 1I11 1111 115 115 0 40 80 120 1110 200 240 2110 0 DISTANCE FROM LEFT BANK HEADPIN (fllllt) Appendix Figure 0-6.Cross section of transects 4,4A,5,and 6 at Island Side Channel (adapted from Quane et ale 1985).. )J ~I J J J ~J J ~I )J J J J l I Calibration Calibration data available at the close of 1984 field season were limited to that obtained for a side channel flow of 338 cfs (56,100 cfs mainstem discharge)(Appendix Table D-4).As a result,an IFG-2 model was used to forecast instream hydraulics based on this single·cali- bration flow.The streambed profile,stages of zero flow,and observed and predicted water surface elevations for this study reach are plotted to scale in Appendix Figure 0-7. The original field water surface elevations (WSEL's)were compared to the model predicted WSEL's for the calibration flow of 338 cfs (Appendix Table D-5).At transect lA,the original field WSEL was surveyed at 93.46 feet.In examining the WSEL1s of transects 1 and 2 (93.33 and 93.41 feet in elevation respectively),it was felt that an error in surveying occurred at transect 1A.As a result,the WSEL for this transect was lowered by 0.1 feet to 93.36 feet.For all other tran- sects,the di fference between the fi e 1d WSEL I S and the model pred i cted WSEL's for the calibration flow were 0.05 ft.or less. The two partial transects (lA and 4A)which represent slackwater habitat were extended out to the principal velocity corridor.This corridor is where most of the flow in the channel occurs.In order to complete the data sets for these two partial transects for use in the model,the associated data from transects 1 and 4 were used.At partial transect lA,the velocities were all negative.In order to use this information in the model,these velocities were treated as positive,as it was felt that the direction of the current would not influence the utilization of this area by juvenile salmon.Only 6.5 cfs or about 2%of the water flowed through this section. Verification Based on the first level of verification conducted by EWT&A,the model does an excellent job of simulating hydraulics between 35,000 and 56,000 cfs mainstem discharge (69 and 416 cfs site flow)(Appendix Figure D-8). Above 56,000 cfs,however,the simulated depth and velocity distri- butions begin to deteriorate in quality.As a result,the model simu- lations were rated good between 56,000 and 64,000 cfs (416 and 692 cfs site flow),acceptable between 64,000 and 70,000 cfs (692 and 984 cfs site flow),and unacceptable above 70,000 cfs mainstem.Below 35,000 cfs mainstem,the site flow was less than 5 cfs,and the model does not simulate accurately below 5 cfs. The velocity profiles produced by the IFG-2 hydraulic model for the two flows,338 and 520 cfs,are compared to their associated observed velocities at two transects (Appendix Figures 0-9 &0-10).The observed and predicted velocities are in good agreement for both flows at tran- sect 1.At transect 6 there is also good agreement between the observed and predicted velocities at the 338 cfs flow.But at the 520 cfs flow, from 85 to 140 feet,there is notable differences between the observed. and predicted values. D-19 ISLAND SIDE CHANNEL Thol"'IV Proflll with Obllrvld and Predlctld Watlr Surfaci ProW .. ~ ....t (Afprol.k".~ • - Thalw ••G.adl.nl'11I.e f••l/mll. Oll ••n.d waler Surfac.EI••a!l"n SI",ulal.d Wal.r Surfoc.El ••allon E~lrapolaled W"ler Surface EI ••ollon EI ••allan ai Zero Fl"w. Thalweg Profile ,..... ,"••••Of e_. &0+00..... _._....._._._._._._._._._._._._._._.-.-._._._.-._._._._._._._._._._._._._,_,_,_,_,_'e."'] •...I.'••SS8 c'.(....pOlo'io..r'''W•." _._._._._._._._._._._._._._._._._...:._._._._._._._._._._._.._._._._._._._._._._._..e.of.••••••11....." ._---:;:%-------------------------------------~ %~ "'0100 i..iii iii iii iii I II • , ,i , ,i , ,iii ii'iii iii Iii iii i ,j u iii i , •iii iii ,. II Cl I N j ,. 0 '"II 0;::c>II..........>,. ~.... IE .. STllEloMBEO STATION I'.." Appendix Figure 0-7.Compa~ison of observed and predicted water surface profiles from calibrated model and surveyed thalweg profile at Island Side Channel (adapted from Quane et al. 1985). J 1 J I t J I I J ,J }.1 I J J J ..... ,- Appendix Table 0-5.Comparison of field measured and model predicted water surface elevations at the cal ibration flow of 338 cfs for Island Side Channel . Transect Water Surface Elevation (ft) Field Model Predicted Difference 1 1A 2 3 4 4A 5 6 93.33 a93.46 93.41 93.44 93.48 93.52 93.56 93.55 93.33 93.36 93.36 93.40 93.46 93.50 93.53 93.56 0.00 0.05 0.04 0.02 0.02 0.03 0.01 - - a Water surface elevation reduced by 0.1 feet to 93.36 feet. 0-21 Application Range of the at .Island Cali brated Hydraulic Mode I Side Channel RM (63.2) Site Specific Flow,cfs o 8 38 115 211 545 984 1283 I I I 1 n ~ 1 I I Io102030405060 10 15 o I N N Mainstem Discharge at Sunshine Station,cfs x 1000 _Excellent _Acceptable II Good D Unacceptable Appendix Figure D-8.Application range of the calibrated hydraulic model at Island Side Channel. J ~i J J )},J J )_I )t ,1 I.il )))J })1 )1 )1 1 -l ISLAND SIDE CHANNEL J Transec t 1 ' 3 ..Observed velocU tes (515 cIs) •Observed veloct Hes (338 cIs) .6.Predtcted veloct t tes (520 cIs) D Predtcted veloct t tes (338 cIs) -u OJ V'l................2Cot-- >-f-- .......-t (J CJ 0 I -IN W W:> 1 .. o-~f·90 1110 1~0 .1~OH_170 'I-I DISTANCE FROM LEFT BANK HEADPIN (ftl Appendix Figure 0-9.Comparison of observed and predicted velocities from the IFG-2 hydraulic model at Island Side Channel,using two flows at the transect 1 discharge site. ISLAND S I DE CHANNEL}.Transec t 6 5 i i ....Observed velocl Hes (543 cfs) •Observed velod t tes (338 cfs) .6.Predlcted ve lOel Hes (520 cfs) [J Predkted veloct t tes (338 cfs) 4 0'ee,6,,,,I ,,,1 1 -U OJ lfI........3--"--- >-I- t--t L}20C) I N ---.J .j:::o LU:> Appendix Figure 0-10.Comparison of observed and predicted velocities from the IFG-2 hydraulic model at Island Side Channel,using two flows at the transect 6 discharge site. J I ).t J I J ,J ))J J el )o~I J - - Application For habitat simulation modelling purposes,the hydraulic simulation model developed for Island Side Channel can simulate channel flows in the mainstem discharge range of 35,000 to 70,000 cfs. 0-25 Mainstem West Bank Side Channel (RM 74.4) Site Description Mainstem West Bank Side Channel is located on the west bank of the main channel Susitna River at river mile 74.4 (Appendix Figure 0-12).It is approximately 2.2 miles in length.The mouth and two heads of this side channel connect directly with the Susitna River. The IFG modelling site in the lower portion of this side channel was 930 feet long (Appendix Figure 0-11).The study site is confined on the west by a steep bank and on the east by a well vegetated island.The portion of the side channel upstream of the study site is separated from the mainstem by a network of side channels and well vegetated islands. A minor channel is located within the study site on the east bank of the side channel.During nonbreached conditions,the side channel primarily consists of a series of pools and small riffles.Groundwater provides the major contribution of flow prior to breaching of the head (Quane et al.1985). The two heads are both located approximately 1.5 miles upstream of the study site (Quane et ale 1985).Breaching of Mainstem West Bank Side Channel occurs when the mainstem overtops either of the two side channel heads.The side channel has been estimated to be initially breached at a mainstem discharge of 19,000 cfs (Quane et ale 1985). Based on a review by Quane et ale (1985)of the stage versus mainstem discharge rating curve (Appendix Figure 0-13),it has been determined that at mainstem di scharges greater than 19,600 cfs,the hydraul i cs within this side channel are directly controlled by mainstem discharge. The site flow that occurs at 19,600 cfs was measured to be 5.7 cfs. Hydraulic information was gathered from five transects (1,2,3,3A,4) in the main channel and three transects (2A,3 in part,3B)in a minor side channel of this study site (Appendix Figure 0-12).The corre- sponding cross sections are presented in Appendix Figure 0-14 &0-15. The two lower transects (1 &2)bisect primarily pool and run habitat, the banks are gently sloping on both sides.On the upper three tran- sects (3,3A,&4)the left bank consisted of an erosional bank and was primarily bordered by alder.For modelling purposes,transects 3 and 3A were ended on a finger-like gravel bar on the right bank which longitu- dinally bisected the site with the main channel on the left and a minor channel on the right which was free flowing at high flows,backwater at median flows,and dry at low flows.This bar began downstream from transect 4 and ended between transects 2 and 3.Transect 3A was placed in order to obtain a better representation of the slow water debris- strewn habitat along the left bank.The main channel habitat of these three transects (3,3A,&4)consisted of run and riffle habitat. Substrate at this site primarily consisted of rubble and cobble.The thalweg gradient of the side channel is approximately 12.3 ft/mile (Quane et al.1985). D-26 ""'" )·····-1 )'1 1 1 1 ..)1 .,1 1 Cl I N -..) I ,~~...:+t Appendix Figure 0-11.Overview of Mainstem West Bank Side Channel (RM 74.4). C1 I N ex> ~'" /),Staff Gage Transect o 260 I I FEET (Approximate Scale) 'r~~~ Appendix Figure 0-12.location of Mainstem West Bank Side Channel study site (RM 74.4). )J J J )),J .)1 J ))I 'I ! 0•••10·Z••7)9 t",a..10''.0011_ .t •0." Q=5.7 ch on lIllI; I~ 3t-,r-----.....--...L;:...-------- WSEL (+90 feet) MAINSTEM WEST BAN K SIC TR 1 GAGE 74.451 3:o ..Ju.1 ow ct::> CIJ <Cw ::aE 0••'10·,·J70S 'las 2.oem .2 •0." U.OOO 1 0 1 7'.700 .,. lI$l\.•10·1•0617 o.0.J9lt •10 ...z 11'04. ~-./COIITlllLl£O ~...U.JCIIlS 01 n.ooo.1s ...VSCL.10·t.lltO 00.Its1 •to .•z •0.16 COIIYIllLLIP 1••IOt'S ,1 It.JOG .•'' t-..I_nOlI Oua.OPfO OCT C_L(O I',ran S ~~11.1110 ." ""[".."t"<1I(.ILOOfO WSEL=92.85 ~.~-MAINSTEM DISCHARGE,SUNSHINE (xlOOO cfsJ f -I~f ~I if"II 1 - ~ Q) .2 0 0).=. ..Jw &-t {/) 3: - -'0-~__._ MAINSTEM DISCHARGE,SUNSHINE ex 1000 cfs) Appendix Figure 0-13.Compari son of rat;n9 curves for Mainstem West Bank Side Channel transect l(Q site)(from Quane et.a 1.1985). 0-29 .....-...__._--_.------------ 105 105 104 CROSS SECTION I 104 ~CROSS SECTION 2 103 STATION 0+00 103 STATION 1+66-102 -102 .ii II 101 II 101II......100 -100 1111 Z 1111 Z liB 2 liB 0 117 I-117i=118 ~118ot>1111 \oJ 1111 \oJ 114 oJ 114oJ\oJ\oJ U 113 \oJ 82 \oJ 112 >III >III 5 110 !;i 110 oJ BII oJ BII W BB W BB It:B7 It:B7 B8 B8 B5 B5 0 100 200 300 400 0 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet I END POINT CROSS SECTION 3 STATION 5 +08 400300200100 450.11=====310.10 "v'«;.11\.....:;/ 1011 104 0 103 I -102 v) - a II 101II...100 1111 Z liB011751111 >1111\oJ 114oJ W 113 W 112 >III i=110otBIIoJBBW It:B7 Be BII 0 DISTANCE FROM LEFT BANK HEADPIN (feet) Appendix Figure D-14.Cross secti on of transects 1,2,and 3 at f~ainstem West Bank Side Channel (adapted from Quane et ale 1985). J J t J J I -I J ]t ,J I »J .J.- )-]»j }1 J 1 ))1 1 -)-1 1 10:1 10:1 104 CROSS SECTION 3A 104 ]CROSS SECTION 4 103 STATION 5 +62 103 STATION 9+32..102 -102".. <II 101 "101-Q) 100 I~-100 ~ Z SS III 0 98 Z 1180t-117 ~j:97«98 98>450 eft ~95I&J 95 310 cft -'94 l,/I&J 114I&J 6 cl.-'93 W 93 W 92 W 92>lit >III t-90 t-110« -'89 «89 I&J 88 -'88I&J0::87 0::87 88 88 85 ,.85 0 too 200 300 400 0 100 200 300 400 0 I DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet)w....... Appendix Figure 0-15.Cross section of transects 3A and 4 at Mainstem West Bank Side Channel (adapted from Quane et al.1985). were this The Calibration Hydraulic data were collected for model calibration at three site flows: 6,310,and 450 cfs,the correspondi ng mean daily di scha rges for the Susitna River were 19,600 cfs,30,500 cfs,and 32,000 cfs,respectively (Appendix Table 0-4).Based on these data,an IFG-4 model was used to forecast ·instream hydraulics.The streambed profile,stage of zero flow,and observed and predicted water surface elevations for the study reach are plotted to scale in Appendix Figure 0-16.All three data sets were used to predict hydraulic information for side channel flows of 6 to 2,431 cfs (mainstem discharges of 18,000 to 75,000 cfs). To evaluate the performance of the hydraulic model,observed and pre- dicted water surface elevations,discharges,and velocity adjustment factors were compared (Appendix Table 0-6).The 15 sets of observed and predicted WSEL's for the five transects of the 3 calibration flows were all within ±0.02 ft.of each other except for 2 sets which were within ±0.10 feet of each other.All the observed and predicted discharges were within 10%of each other and all velocity adjustment factors were within the good range of 0.9 to 1.1.Additionally,the stage infor- mation of the model was compared to available rating curves (Appendix Figure 0-13). Transect (3A)was pl aced about 60 feet upstream from transect 3 to represent the slackwater debris area along the left bank of the upper portion of this study site.In order to complete this data set for transect 3A for use in the model,the velocity information from transect 3 for the two site flows of 310 and 450 cfs were incorporated into transect 3A cross sectional area and water surface elevations.After incorporating this information into transect 3A,the discharge for the 310 cfs site flow,however,did not fall within 10%of the respective discharge that was calculated at the discharge transect.As a result, velocities for the 310 cfs site flow were adjusted upward by 17%. At the low flow measurement of 6 cfs,the ve loci ty measu rements made compl etely across transect 3A.The di scharge cal cul ated at site was 18%higher than calculated at the discharge transect. velocities at this transect were therefore reduced by 15%. At transect 4 the water surface elevations were not similar across the transect at the 6 cfs flow measurement.Therefore,a weighted average water surface elevation was calculated for this transect. At higher site flows several small side channel/backwater areas existed which were not represented in the IFG-4 analysis.In order to evaluate this potential habitat several transects were placed across one of these areas,weighted usable area was to be determined by hand calculations. However,this was not done because it was determined that this side channel habitat was so small compared to the total area being hydrau- lically modelled that it would not affect the total weighted usable area response. 0-32 -- - ""'" ,~ - - j »)1 )1 ]-J }1 ))l'j )J MAINSTEM WEST BANK SIDE CHANNEL Thalwe"Profile with Observed and Predlc;ted Water Surface ProfIJ e. OOV~fi'////W "-AJr.;;; EAST BANK MINOR CHANNEL ".. ..Thal••,Su,..,Dolo-841010 TIlal••,G.adi.nl-12.1I I••llmil. Ob •••••d Wal.,So.Ia..EI••allon SlllIulal.d Wal.,50,10••El ••olion Exltapolohd Wal.,Su"o••EI ••ollon EI ••allon of Z.,o Flo. Thai••,P,ofll.w# / 9 2~O ~ (AwOl 5eol.) M41NSTEM WEST lANK SIDE CHANNEL TIlANSEn u.TA""ntt J TAAHn:CT 58 I...1811 ,...---tU'i'i ,iii ii'iii'iii iii ~+()O 10+00 STREAMBED STATION (100'I ""'_=m~;;Ml'9'#-%-...,.,/;/~ AOftO OAGEl 74.4"--1 la"NUn I IAO'-.'G"". 14.411 T".NIICT I lIllANI!IECT''.....'lel IA IAOf 60 GAG.£.11•."154 __ 1T••"lun. STREAMBED STATION 11001' Appendix Figure 0-16.Comparison of observed and predicted water surface profiles from calibrated model and surveyed thalweg at Mainstem West Bank Side Channel (adapted from Quane et ale 1985). Appendix Table 0-6.Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Mainstem West Bank side channel hydraulic - model. Strea'mbed Water Surface Station Elevation Oi scharge Velocity Observed Predicted Observed Predicted Adjustment """"(ft){ft}(ft)(cfs)(cfs)Factor ~ 0+00 92.85 92.86 6.0 6.3 1.005 1+66 92.86 92.87 6.9 7.2 .991 5+08 93.25 93.26 6.9 7.2 1.004 5+62 93.51 93.52 5.8 6.1 .996 9+32 95.06 95.06 5.1 5.4 1.013 Qo =~Qp =~...... I 0+00 94.62 94.61 312.8 315.7 1.030 1+66 94.64 94.64 301.3 307.5 1.024 5+08 94.85 94.86 306.4 318.2 1.007 -5+62 94.93 94.99 292.8 288.6 .993 Qo =301.0 Qp =308.0 ~ 0+00 94.97 94.98 460.4 457.0 .974 1+66 95.00 95.00 446.1 438.2 .975 5+08 95.19 95.18 470.6 455.2 .994 5+62 95.29 95.23 409.6 415.3 1.001 - 9+32 96.54 96.45 473.9 451.9 .969 00 =452.0 Qp =444.0 Qo is the mean observed calibration discharge. ~, Qp is the mean predicted calibration discharge. """" - D-34 1'fIlI!!lIiI\ - - ..- """ Verification Based on the first level of verification by EWT&A,the model does an excellent job of simulating channel hydraulics between 18,000 and 21,000 cfs mainstem discharge (6 and 20 cfs site flow)(Appendix FigureD-17). Above 21;000 cfs,simulated water surface profiles deviate somewhat from field observations.As a result,the model was rated good between 21,000 and 28,000 cfs mainstem discharge (20 and 200 cfs site flow),and between 28,000 and 34,000 cfs mainstem discharge (200 and 500 cfs site flow)the model again was rated excellent.Two calibration data sets were collected within this range.Above 34,000 cfs,the quality of the hydraulic simulations begins to deteriorate as the slope of the site flow versus WSEL relationship flattens as a result of channel geometry. The deviation between the regression line developed within the model and that of the rating curve developed independently for the site increases with discharge until the model simulations are no longer acceptable. The model simulations were rated good between 34,000 and 41,000 cfs (500 and 727 cfs site flow),acceptable between 41,000 and 48,000 cfs (727 and 1000 cfs site flow),and unacceptable above 48,000 cfs mainstem dis- cha rge. At the second level of verification there is good agreement between the predicted and observed values of depth and velocity (Appendix Figure 0-18).At the higher velocities (>2.5 ft/sec)they begin to spread apart though.In Appendix Table D-7 the results of the statistical tests are shown.There is again good agreement shown between the observed and predicted values for both velocity and depth.The index of agreement (d)is almost one,the total root mean square error (RMSE)is largely composed of the unsystematic RMSE,and the y-intercept (a)is close to zero with a slope (b)of almost one.? Application For habitat simulation modelling purposes,the hydraulic simulation model developed for Mainstem West Bank Side Channel can simulate channel flows in the mainstemdischarge range of 18,000 to 48,000 cfs. 0-35 Application Range of the Cali bra ted Hydraul ic Mode I at Mainstem West Bank RM (74.4) Site Specific Flow,efs o 13 307 690 1080 1555 2118 2431 I I ~I I I II III ,. ~H ,'\',,\, I I ,I Io10203040506070 75 CI I W 0'\MainstemDischarge at Sunshine Station,cts x 1000 _Excellent _Acceptable •Good D Unacceptable Appendix Figure D-17.Application range of the calibrated hydraulic model at Mainstem West Bank Side Channel. )J ),.J t )J J I ,9 )_J !).J )I 1 ))1 )1 ]J 1 1 -J )) MAINSTEM WEST BANK SIDE CHANNEL MAINSTEM WEST BANK SIDE CHA~~NEL 4- lio o~~ rP S~ rdb;.0 a o~C Cb~a o 0 a a a 0 n..EI 0 0 (jlIo 0 ~'tl rP f 0 8 2 3 OBSERVED VELOCITY (IT/SEC) Observed v.s.Predicted VelocitiesObservedV.s.Predicted Depths 4-I I 4- 00 a 0 3.5 -l oettt I 3.5 fO~'" 3 0 3Iiw,....III t t'-' I:2.5 '-'2.5 I-~0- W i3a2g 2a~w t>aa1.5 w 1.5 w l- n:~ 0-a 1 w CJ n: I Q. w.....0:111:I 0.5 0III I I I 0 1 2 :3 4-0 OBSERVED DEPTH (IT) Appendix Figure D-18.Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraulic model at Mainstem West Bank Side Channel. ·~.J !J J ,J J )cl )I J I ~I ):I ill r - .- r- I Circular Side Channel (RM 75.3) Site Description Circular Side Channel is located on the west bank of the Susitna River at river mile 75.3 (Appendix Figure 0-19).It is approximately 0.9 .miles long and is separated from the mainstem by a large well vegetated island.An extensive backwater area occurs in the lower portion of the study site.A network of small channels at the head provide mainstem flow into the site after breaching.Prior to breaching,flow is greatly reduced and the channel is composed of 1a rge pools connected by small riffles (Quane et a1.1985).' Breaching of Ci rcu1ar Si de Channel has been estimated to occur at a mainstem discharge of 36,000 cfs (Quane et a1.1985).It has been determined that the hydraulics within this side channel are governed by mainstem discharge at mainstem discharges exceeding 36,000 cfs.The site flow that occurs at this mainstem discharge is estimated to be 26.8 cfs (Appendix Figure 0-20)(Quane et a1.1985). Based on assessments by Quane et a 1.(1985),backwater does not occur during non-breachi ng mai nstem di scharges.At breaching rna i nstem.di s- charges of 55,200 to 66,700 cfs,however,an area of backwater was found to occur upstream to a point approximately 90 feet above transect 2A. At a mainstem discharge of 42,500cfs,backwater has been determined to extend slightly past transect 2. The IFG modelling study site in the upper half of Circular Side Channel is 820 feet (Appendix Figure 0-21).The thalweg gradient of this study site is 14.3 ft/mi1e (Quane et aL 1985).Riparian vegetation along both banks consists mostly of alder and cottonwood.Substrate within the lower reaches of the Circular Side Channel site consisted predomi- nately of silts,sands,and gravels changing to rubbles at the upper reaches.Hydraulic information .was gathered from six transects estab- lished at this study site (Appendix Figure 0-21).The channel is relatively straight and the cross sections are generally box shaped in confi gurati on (Appendi x Fi gures 0-22 &0-23).Transects 1 and 2 were located in shallow backwater.Transect 2A was located in a transitional area which became run habitat at higher flows.Transect 3 was located in riffle habitat.Transect 4 was located in run habitat at the end of a pool,transect 5 bisected this pool. Calibration Hydraulic data were collected at two calibration flows:50 and 204 cfs (Appendix Table 0-4).Mean daily discharges for the Susitna River on the dates that calibration data were collected at the Circular Side Channel study site were 42,500 and 55,200 cfs.An IFG-4 model was used to forecast instream hydraulics based on these two calibration flows. The streambed profile,stages of zero flow,and observed and predicted water surface elevations for the study reach are plotted to scale in Appendix Figure 0-24.The two data sets were used to predict hydraulic information from side channel flows of 6 to 733 cfs (mainstem discharges of 25,500 to 75,000 cfs).. 0-39 C), .p-o I .,,~..~j,t Appendix Figure 0-19.Overview of Circular Side Channel (RM 75.3). J l J ~J J j J J )")]t J j J ) -! - - ..... o:J) (j...., :s:. O'...... u. 0.wa: ::len«w:a: ~ O,c •10~.'.1064 0..4.lOIi5 ..2 •a.." ! ..... <II (j...., :: 0..... LI.. 0.wa: ::len« UJ :E • '•.•~",',_.0''.0'/ ,1 ..o~./ 0=26.8 cis •--------t I ~ I~tg l~ C1RCULAR SIDE CHANNEL TR4 GAGE 75.354 ........_.i'-------------""'----....,...;.. MAINSTEM DISCHARGE,SUNSHINE (x 1000 Of5)WSEL (+85 feet),. ~Q=36,000 cll Y ',,""'O<LEG J6 ..00CI ~Q.$.".100 C-'\ WSEL=90.15 I llS/:L •10'1.1619 QO.<55:•85 - -..lvO~"~1IQt COllY_LEO ",JOO ~o!16.000 cr, lISE<•100·110J QO.lIlSS •os ,.:t •0.'1....I Wen:s: "il __ MAINSTEM DISCHARGE.SUNSHINE (x 1000 cfsJ Appendix Figure 0-20.Compari son of rating curves for Ci rcul ar Si de Channel Transect 4 (from Quane et.al.1985). D-41 CJ I +:> N Appendix Figure 0-21.Location of Circular Side Channel study site (RM 75.3). 'uk,.". ,£It;__ U'~', ~(;,q. q.'~:r,,,-\y~t-',u""" ))1 ~J l 1 }J J l ]J t ])) ]J J j 1 J ]1 1 1 )J }))1 CROSS SECTION I STATION OtOO 'c ~204 ell ~.-----J !So 010 CROSS SECTION 2 STATION H98 , 2110240200 I 1601208040 105 104 103..102....101-100..... Z 99 0 1111 ~1J7«1J6>95I&J ..J 1J4I&J 93 I&J 1J2>IJ1 I-IJO«88..J I&J 1111It:117 1111 1I11 021102402001801208040 105 104 103....102....101~.....100 Z 99 0 911 ~97 ~98 I&J 1J5 ..J 1J4I&J 93 I&J 1J2>1J1 I-110«119..J I&J 1111 It:117 1111 115 0 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet) Cl I :~:1CROSS SECTION 2A-Pow ~103 STATION 2+65..102GO..101-.....100 Z 88 0 88~87«86> 1&.1 85 ..J 841&.1 83 W 1J2>IJ1t=«80 ..J 88 I&J 8110:87 811 115 0 40 80 120 1110 200 240 2110 DISTANCE FROM LEFT BANK HEADPIN (feet) Appendix Figure D-22.Cross section of transects 1,2,and 2A at Circular Side Channel (adapted from Quane et al.1985). DISTANCE FROM LEFT BANK HEADPIN (feet) 2802402001601208040 --,..,204 cr..........~50c"... CROSS SECTION 4 STATION 6+63 lOS lOS 104 CROSS SECTION 3 104 103 STATION 4+33 1-03....102 ....102GlGl..101 Gl 101...... 100 ~100 Z 99 Z 99098098 I-97 I-97et«98>9a >ILl 9S ILl 9S..J 94 .J 94ILlILl 93 93 ILl 92 ILl 92>>j:91 204 c"~91 et 90 50 cfa 90 ..J 89 ..J 89 ILl 88 ILl 880::0::117 87 116 86 liS liS 0 40 110 120 180 200 240 2110 0 DISTANCE FROM LEFT BANK HEADPIN (feet) CROSS SECTION 5 STATION 8+20 lOS 104 CJ 103 I -102 ~ .... ~ Gl 101Gl...100 99 Z 98097j:98~9SILl94..J ILl 93 ILl 92 >91 5 90 89.J 118ILl 0::117 118 liS 0 40 110 120 180 200 240 2110 DISTANCE FROM LEFT BANK HEADPIN (feetl Appendix Figure 0-23.Cross.section of transects 3,4,-and 5 at Circular Side Channel. I I )J J "' J J j I -]I J J •_.~_J J )1 'I )1 1 1 )~i 1 J )1 ] CIRCULAR SiDE CHANNEL Thaiwill Proflll wllh Ob.erVld and Prldlclld Water SurfaCI Pro'''" .~/ ~ tAW::,·~~I.1 o WI/) Tholweo Grodl.nt·14.1 'ti.llmlle Ob ....ed Wate,Su,'a..Ele.atlon Simulated Wat.,Su,'a••Ele.ol'on EIl,apalaUd Wale,Su,'o.e EI ••allon Elavollon 0'Ze,o Flow Thol.eg P,oflle 50+00u+oo10"'001&....00 t"ANSEcT S 10+00 ,.AN,leT 4 "'..filIHCT I I-811..-••1 -I '''ANlE'eT -I w otoo t • .. a .. I ;; ..j:::o ;; U1 Z H 0;::..>t.W ..J W W::-tf;::.. ..J W 0:o• .. ro 0';'00 STREAMBED STATION II..,) Appendix Figure 0-24.Comparison of observed and predicted water surface profiles from calibrated model and surveyed thalweg profile at Circular Side Channel (adapted from Quane et al. 1985). To evaluate the performance of the hydraul ic model,observed and pre- di cted water surface el evati ons,di scharges,and vel oci ty adjustment factors were compared (Appendix Table 0-8).Because of the 2 cali- bration flows only a 2 point rating curve was formulated.In evaluating the performance of the model,observed and predicted WSELls and dis- charges were the same because of this rating curve.Velocity adjustment facto rs we re a11 with in the good ra nge of 0.9 to 1.1.Add iti ona 11 y,.the stage information of the model was compared to the rating curves estab- lished by Quane et al.1985 (Appendix Figure 0-20). At the high flow measurement of 204 cfs,the original field measured discharge at transect 2 was 34%lower than that calculated at the discharge transect.In order to use this information in the model,the individual velocity measurements were all adjusted upwards by 52%.Why there was such a large discrepancy between flows at this particular transect when the four other transect flow measurements were within 9% of the discharge transect measurement is unknown. At transect 5 there was a change in the channel cross section from when the actual cross 'section survey was done and when the two calibration flows were made.Between the cross section survey of September 5,1985, and the two calibration flow measurements July 24 and August 17,1984,a flood event occurred on August 26,1984.After this flood,the right side of the channel at transect 5 was scoured out.In order to avoid violating one of the underlying assumptions of the model,(i .e.,that a rigid stream channel exists)the cross section determined from the two calibration flows was used in the model. During the 50 cfs calibration flow measurement a water surface elevation was not surveyed for transect 5.In order to obtain a water surface elevation for the model,a value was calculated from the average of the depth measurements added to the corresponding cross section elevations of the 50 cfs flow measurement. Verification Based on the first level of verification by EWT&A,the model does an excellent job of simulating channel hydraulics between 39,000 and 57,000 cfs,mainstem discharge (38 and 213 cfs site flow).Above 57,000 cfs, the simulated depth and velocity distributions begin to deteriorate in quality.The model simulations were therefore rated good between 57,000 and 60,000 cfs (213 and 268 cfs site flow),acceptable between 60,000 and 63,000 cfs (268 and 334 cfs site flow),and unacceptable above 63,000 cfs mainstem discharge.Below 39,000 cfs,the model simulations were also rated less than excellent as forecasted velocity and depth distributions deteriorated in quality.The model simulations were rated good between 36,000 and 39,000 cfs mainstem discharge (27 and 38 cfs site flow)(Appendix Figure 0-25).Below 36,000 cfs mainstem (con- trolling discharge),insufficient information is available to evaluate the model. At the second level of verification there is excellent agreement between the observed and predicted velocities and good agreement between the 0-46 ~, - - - - Appendix Table 0-8.Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Circular Side Channel hydraulic model. .... Streambed Wa ter Su rface Station Elevation Discharge Velocity Observed Predicted Observed Predicted Adjustment (ft)(ft) (ft)(cfs) (cfs)Factor 0+00 89.28 89.28 44.4 44.4 1.000 1+98 89.30 89.30 47.9 47.9 .998 2+65 89.41 89.41 56.0 56.0 1.000 4+33 90.20 90.20 43.7 43.7 1.000 6+63 90.60 90.60 50.9 50.9 .997 8+20 90.62 90.63 53.6 53.6 1.000 Qo =49.0 Qp =49.0 0+00 90.29 90.29 202.8 202.8 .998 1+98 90.27 90.27 203.1 203.1 .987 2+65 90.31 90.31 198.4 198.4 .999 4+33 90.66 90.66 176.9 176.9 .998 6+63 91.29 91.29 199.9 199.9 1.000 8+20 91.32 91.32 194.2 194.2 1.000 00 =196.0 Qp =196.0 Qo is the mean observed calibration discharge. Qp is the mean predicted calibration discharage. 0-47 Application Range of the Calibrated Hydraulic Model at Circular Side Channel RM (75.3) Site Specific Flow,cfs o o 10 2 20 12 30 43 40 118 50 268 60 537 733 70 75 Cl I ~ 00 Mainstem Discharge at Sunshine Station I cfs x 1000 mI Excellent II Good -Acceptable y Unacceptable Appendix Figure 0-25.Appli~ation range of the calibrated hydraulic model at Circular Side Channel. _I !J I ],I J J .1 J J 1 I I I - .- - observed and predicted depths (Appendix Figure 0-26).The results of the statistical tests also indicate good agreement between the predicted and observed values for both velocity and depth (Appendix Table 0-7). The index of agreement is near one,the total RMSE is mostly composed of the unsystematic RMSE,and the y-intercept is close to zero with a slope of almost one. Application For habitat simulation modell ing purposes,the hydraul ic simulation model developed for Circular Side Channel can simulate channel flows in the mainstem discharge range of 36,000 to 63,000 cfs . 0-49 CIRCULAR SIDE CH.ANNEL CIRCULAR SIDE CHAt'-JNEL Observed vs.Predicted Depths Observed vs.Perdieted Velocities 4.I 4 (] (]0 3.5 -l (](]3.5 ~Da (]~...rP 3 u 3 (] " (]~D~BtP W tJ (I)cf"w t tp/LL '--'2.5 Ire ~£J 'V 2.5 :r o aD ~(][Jl-I!D-ow20 20..J 0 W , W > I-001.5 w 1.5 0 I- w ·2 It 0 D-1 wC1ItID-Ol C>(] 0.5 o pw=I oV (] I I I I I I I 0 1 2 :5 4 0 1 2 3 4 OBSERVED DEPTH (FEET)OBSERVED VELOCITY (FT/SEC) Appendix Figure 0-26.Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraulic model at Circular Side Channel. ,1 1 I _J J '1 )l I J )J I })I - ,..... - ,..... ! ,- ..... I Sauna Side Channel (RM 79.8) Site Description Sauna Side Channel is located on the west bank of the Susitna River at river mile 79.8 (Appendix Figure D-27).It is approximately 0.2 miles long.Both the mouth and head of the side channel are connected to a larger side channel of the mainstem Susitna River.For the most part, the side channel is confined on the west side by a high bank ~nd on the east by a large sparsely vegetated gravel bar.A smaller side channel enters just below the head of Sauna Side Channel on its west bank.This side channel conducts flow to the study site during high mainstem discharges,but dewaters before the head of Sauna Side Channel becomes unbreached.Breaching flows result from overtopping of the side channel that adjoins the head on the east bank of Sauna Side Channel.Prior to breaching,the channel is composed of two large interconnected pools whose water levels are maintained from ground water seepage originating from the vicinity of the head.An extensive log jam at the head of Sauna Side Channel influences the flow into this side channel. Based on field observations and stage/discharge relationships,the mainstem discharge estimated to initially breach Sauna Side Channel was 37,000cfs (Quane et ale 1985).A controlling discharge of 38,000 cfs was determined for this side channel also based on this stage/discharge relationship.A side channel flow of22.5 cfs was estimated to occur at the 38,000 cfs mainstem discharge .as derived from the stage versus streamflow rating curve (Appendix Figure 0-28).Quane et ale (1985) determined that backwater does not occur in Sauna Side Channel during non-breaching mainstem discharges.During breaching discharges of 54,600 to 56,700 cfs,however,backwater was observed to occur through- out the Sauna Side Channel study site. The IFG modelling site,located approximately 2,000 feet from the mouth of this side channel,was 480 feet long (Appendix Figure 0-29).The thalweg gradient at this site is 10.4 ft/mile (Quane et ale 1985). Substrates throughout this site consist primarily of sands and silts. The water is slow moving with velocities usually less than 1.0 ft/sec. The left bank at the site is an erosional bank with a height exceeding fi ve feet;ri pari an vegetati on along thi s bank consi sts of alder and birch.In contrast;the left bank isa depositional bank with no riparian vegetation. Four transects were located within this study site (Appendix Figure 0-30).Transects 1 and 2 were located in shallow pool habitat whereas transects 3 and 4 were located in deeper pools. Calibration Hydraulic data were collected at a calibration flow of 52 cfs corre- sponding to a mainstem discharge of 52,000 cfs (Appendix Table 0-4). Based on this single calibration flow,an IFG-2 model was used to forecast instream hydraulics of this study site.The streambed profile, stage of zero flow,and observed and predicted water surface elevations for the study reach are plotted in Appendix Figure 0-31.This data set D-51 FEET (Approximate Scale) - - RM79Ef) 500 ! River MUe a ! Appendix Figure 0-27.Overview of Sauna Side Channel (RM 79.8). 0-52 ! ..... <Il :3 3:o ..J I.L. o W0:. :::>en«w ::!i: !! 0'IC III 10-••1214 0...Z.:."" ,.!011 0.10 3:o ..J I.L.C o W 0::::>en«w ::!i: 0...166 •11"11l5lL •lSI t._ "I •o.•a Q =22.5 ctl SAUNA SIDE CHANNEL TR2 GAGE 79.8S2 COIT1IGll [~ 1I.GOCI ~0 S 17 .100 cf' "'0.•1~·1.7_O·uSt •IS .t 6.11 lIlT.a-r....lEl 15._~6 ~la.a c" '"'£l1IJI.T 1111 1100.00£6 -;,._....,,10-.1.-. MAINSTEM DISCHARGE,SUNSHINE (x 1000 cfsJ WSEL ("85 feet) It + I I -- - -~03 _,-m MAlNSTEM DISCHARGE.SUNSHINE (x 1000 cfsJ Appendix Figure 0-28.Comparison of rating curves from Sauna Side Channel transect 2 (from Quane et.at.1985). 0-53 Appendix Figure 0-29.Location of Sauna Side Channel study site (RM 79.8). 0-54 -, -~J 1 }J ))})}1 )-J ]-}J » r----. 20 40 150 110 100 120 140 160 1110 200 CROSS SECTION 2 STATION It81 '\,'-&2 Of. 10~ 104-10;' 011 102......101- Z 100 ~1111 I-lUI CI 117 >1115 ILl .J II!! ILl 114 ILl 113 >92 I-91 <r 110 .J 1111W 0::811 117 118 II!! 02040601101001201401110 1110 200 CROSS SECTION I STATION 0+00 10~ 104 103-102;101...-100 99 Z 11110117~1111 II!!I.Ll 114.J I.Ll 113 ILl 112 >111 S 110 1111.J 1111ILl 0::117 116 II~ 0 DISTANCE FROM LEFT BANK HEADPIN (feeO DiSTANCE FROM LEfT BANK HEADPIN (feet I 0 I 10!!10~ U1 104 CROSS SECTION 3 104U1103STATION3t71103--011 102 ;102..-101 ..101-100 -100-Z 1111 Z 1111211110116 I-117 I-117~118 <r 1115 ILl II!!>II!!.J ILl W 114 .J 114 113 ILl 113 W 112 ILl 112>111 >11152oliI-110 I-110<r .J 1111 <r 1111 ILl .J 0::1511 W 116 117 0::117 118 116 II~II~ 0 20 40 80 110 100 120 140 1110 1110 200 0 DISTANCE FROM LEFT BANK HEADPIN (feet) CROSS SECTION 4 STATION 4-t81 20 40 60 110 100 120 140 160 1110 200 DISTANCE FROM LEFT BANK HEADPIN (feet) Appendix Figure 0-30.Cross section of transects 1,2,3,and 4 at Sauna Side Channel (adapted from Quane et al.1985). SAUNA SIDE CHANNEL Thellwell Profile with Ob..rved ond Predicted Wot.,Sur'oCi Pro'll .. .. • ~ Thai...Sarv.,Dot.·84 100' Thai •••Gracll.nt-10.4 ,..1111111. OII..rv.cI Wal.r Sar'ace EI.vatlon SllIIalat,d Wat.r Sur'ac.EI.vaIlOll Ellropo'at.d wat.r Sur'"c,EI..all"n Elavc"'an ,,'Z.ro F'o. Tho'.,,,Pr,,'lI. .. ••,_._.._._._._..-._._.•._._._._._.._.._.._._._._...._._.-._11--.11:1:lE.ir.p.JoIIOlll ,e••._.._.._.._._._._._._._._.._._._._._._._._._._._._._._0_.-.-lec,..t .,..,.,,110 ••4.1 "-'~---~.~______~I':~~ .. '"II>5 .J '""'lI: CJ t U"l ('7) II ,IIIA"'E"•TIllAMUCT I tu.aU,s JRA.'II;'• I.......10"'00 801 ,iii i i • ; :i ;iii iii i I:,ii'iii I i 81:,ee;iii iii ,i"'},ii,I Ii'i D iii i,'. 0+00 8+~ STREAMBEO STATION IIllil Appendix Figure 0-31.Compari son of observed and predi cted water surface profil es from calibrated model and surveyed thalweg profile at Sauna Side Channel (adapted from Quane et ala 1985). J J }j J !J ]J 1 ].D 1 1 1 - - - was used to predict hydraulic information from side channel flows of 5 to 93 cfs (mainstem di scharges of 21,000 to 75,000 cfs).To eval uate the performance of the hydraul ic model,observed and predicted water surface elevations were compared (Appendix Table 0-9).Additionally, the stage information of the model was compared to the rating curves established by Quane et al.(1985)(Appendix Figure 0-28). It was difficult to calibrate hydraulic'information at this site because very limited field data were available.A site flow versus WSEL rating curve could only be developed for transect 2 (Appendix Figure 0-28). The IFG-2 model is essentially a water surface profile model and a critical variable for calibrating it,is the water surface elevations of simulated flows.Data,however,;s only available for transect 2 and not for any of the other three transects.The actual velocity measure- ments from other measured flows at the discharge transect,however,can be compared to the model predicted velocities for those same flows.At the discha,rge measurement for transect 2,however,there were only two flows that were far enough away from the 52 cfs measurement to be usable (38 and 68 cfs).Thus,the information available to hydraulically calibrate the IFG-2 model for this site consists of the water surface elevations and velocity measurements ,for all four transects at the calibrating flow of 52 cfs,and water surface elevations and velocities for the two other site flows of 38 and 68 cfs at transect 2. This site is influenced by backwater and the effects are more pronounced at the 68 cfs flow.From the field data,the observed top width is greater by 20 feet,the water surface elevation is 0.93 feet higher,and the average velocity is 0.20 ft/sec slower than predicted by the model. At the 38 cfs flow,the effect seems to have reversed,with the observed widths being similar,the WSEL 0.08 feet lower,and the average velocity 0.09 ft/sec faster than predicted by the model (Appendix Table 0-10). In the calibration process,the original field WSEL was reduced by 0.1 feet.This adjustment was made in order to obtain water surface ele- vations that agreed more closely to the lower site flows.It was felt that this adjustment would make the model,in terms of predictability, more sensitive at the lower site flows.By reducing the WSEL of tran- sect 1 by 0.1 feet,the difference between the field and the model WSEL at the 38 cfs flow was reduced from 0.18 feet,when the calibration discharge WSEL was 90.71,to 0.08 feet,when the calibration discharge WSEL was 90.61 feet (Appendix Table 0-10). As a result of a flood on August 26,sediments were deposited in the study site resulting in changes "in all the cross sections derived from the calibration flow on July 23.As a result,the cross sections obtained during the September 15 survey were used in the model until the water's edge of the calibration flow was reached,then the cross sections from the calibration flow were used. When measuring the velocities and depths at each of the transects,the discharge calculated at transect 4 was 16%lower than the 52 cfs site flow calculated at the discharge transect.In order to utilize this information in the model,the velocities were adjusted upwards by 16%. 0-57 Appendix Table 0-9.Comparison of field measured and model predicted water surface elevations at the cal ibration flow of 52 cfs for Sauna Side Channel. *Field water surface elevations were reduced by 0.1 feet. 0-58 ..... - - - ~, - )1 I 1 1 ]--..1 UI<A}T /F-Aut]11) 4/19/85 ANDY/Tables 1 Appendix Table 0-10.The effects of the backwater at Sauna Side Channel,information obtained from transect 2. Top Width (ft) Field Model Average Velocity (ft/sec) Field Model o I tJ'1 t..O Original Modified Site WSEl (ft)WSEl (ft) Flow (cfs)Field Model Field Model 68 91.85 91.06 91.85 90.92 52 a 90.n b 90.74 90.61 c 90.62 38 90.24 90.42 90.24 90.32 a Calibration flow b Original field WSEL input into model c Field WSEl reduced by 0.1 ft 77 .0 53.5 50.5 55.0 53.0 52.0 0.32 0.53 0.51 0.52 0.49 0.42 No stage-site flow rating curve was developed for transect 1.When inputting other flows into the model,the IFG-2 requires either the associated WSEL for this flow or the slope.Because the WSEL could not be obtained for other flows at this transect,a slope value of 0.00005 was input instead.This value was generated by the model from transect 1 at the calibration flow of 52 cfs. Verification The dominant influence of backwater on channel hydraulics makes the site a poor candidate for application of IFG-2 modeling techniques.However, because only one data set was collected,application of the IFG-4 hydraulic model was not possible. Based on the first level of verification by EWT&A,the IFG-2 model for this site does an excellent job of simulating channel hydraulics between 48,000 cfs and 58,000 cfs mainstem discharge (34 to 52 cfs site flow) (Appendix Figure 0-32).Within this range,predicted WSEL's,depths, and velocities are in close agreement with field information (evaluated at 38 cfs by discharge.measurement made by Quane et a 1.(1985).The predictive capability of the model within this range provides evidence that the backwater influence within the study site is lessening with decreasing discharge. Below 48,000 cfs mainstem,there is increasing disagreement between the WSEL IS predicted by the model and those extrapol ated from the rati ng curve.At a 23 cfs site flow,the difference in predicted WSEL between model and rating curve equation has increased to approximately one foot at transects 1 and 2.Although there is evidence that suggests that the model may be a more accurate predictor of WSEL's than the rating curve equations below 48,000 cfs ma;nstem,insufficient information exists to resolve the difference with confidence.Since depths become shallow within this range,predictive errors in WSEL can result in significant errors in predicted depths and velocities.For this reason,the recom- mended extrapolation range ;s limited below 48,000 cfs. Above a 48,000 cfs mainstem discharge,there is increasing,disagreement between the WSEL's predi cted by the model and those observed i·n the field.One of the premises of the hydraulic theory that is the basis of the IFG-2 model is that the water surface profile of the study reach is controlled by its slope.This premise is violated when the water surface profile is influenced by mainstem backwater.From examination of discharge measurements made at 48 and 68 cfs it is apparent that the influence of backwater is increasing with stage above 58,000 cfs mainstem. Overall,the recommended extrapolation range is limited above 58,000 cfs.The model simulations were rated excellent between 48,000 and 58,000 mainstem discharge (34 to 52 cfs site flow).Good between 46,000 and 48,000 (31 to 34 cfs)and from 58,000 to 60,000 cfs (52 to 58 cfs). Acceptable between 44,000 and 46,000 cfs (28 to 31 cfs)and 60,000 to 63,000 cfs (58 to 62 cfs).The model was rated unacceptab 1e below 44,000 cfs and above 63,000 cfs mainstem discharge (Appendix Figure 0-32)• 0-60 - - - )1 --1 1 I )1 J -1 Application Range of the Calibrated Hydraulic Model at Sauna Side Channel RM (79.8) Si te Specific Flow,cfs 5 12 22 37 56 80 93 I I I I I I I==f. I§t I ====~=,I I I I Io10203040506070 75 CI I O'l I-' Mainstem Discharge at Sunshine Station,cfs x 1000 IillI Excellent Iii Good -Acceptable D Unacceptable Appendix Figure 0-32.Application range of the calibrated hydraulic model at Sauna Side Channel. The velocity profiles produced by the IFG-2 model at transect 2 were compared to the observed velocities at flows of 38 and 68 cfs (Appendix Figure 0-33).Because this site is primarily a backwater area and the IFG-2 hydraulic model is not a backwater model it was thought that calibrating the model to more accurately predict at the lower flows woul d be more criti ca 1 than at the hi gher flows.Thus at the 38 cfs flow there is found a better correspondence between the observed and predicted velocities.At the 68 cfs flow the backwater becomes more apparent.A majority of the observed velocities are lower than the predicted velocities and many of these values are lower than individual 38 cfs flow velocities.Because of the overa,.l low velocities,1.0 ft/sec,it was felt that this was the best compromise in applying this model to the Sauna Side Channel site. Application For habitat simulation modelling purposes the hydraulic simulation model developed for Sauna Side Channel can simulate channel flows in the mainstem discharge range of 44,000 to 63,000 cfs. D-62 ~1 - - """ ~! ..,. ).1 .1 ).-1 ]_._))~)-]l-}1 SAUNA SIDE CHANNEL}Transect 2 ..Observed ve lod ties (68 cts) •Observed velod ties (38 cts) A Predicted velocl ties (68 cts) CI Predlcted velod t les (38 cts) ...... o .20 30 40 DISTANCE FROM LEFT BANK HEADPIN (ft) Appendix Figure 0-33.Co~parison of observed and predicted velocities from the IFG-2 hydraulic model at Sauna Side Channel using two flows at the discharge transect. Sunset Side Channel (RM 86.9) Site Description Sunset Side Channel is located on the east bank of the Susitna River at river mile 86.9 (Appendix Figure 0-34).It is approximately 1.1 miles long and is separatea from the main channel of the Susitna River on the west by a network of vegetated islands and side channels.The channel is confined on the east by a high cut bank.Prior to breaching,the side channel is composed of a sequence of pools and riffles.During this period,flow is maintained in the main channel by groundwater seepage and upwelling.After breaching,flows up to 3,900 cfs have been measured (Quane et ale 1985). Breaching of Sunset Side Channel results from the direct overtopping of the head of the side channel by the mainstem Susitna River.Based on assessments by Quane et ale (1985)the side channel initially breached at 31,000 cfs and controlled at a mainstem discharge of 32,000 cfs. The associated site flow at the controlling discharge has been esti- mated to be 45.8 cfs while a flow of 41.1 cfs is derived from the flow versus mainstem discharge rating curve (Appendix Figure 0-35). Based on assessments by Quane et a1.(1985)a backwater area does not occur in this side channel during unbreached conditions.But at breach- ing mainstem discharges ranging from 56,000-66,700 cfs,an area of backwater was observed to extend upstream approximately 1,100 feet to a point between transects 1 and 2. The IFG modelling site within Sunset Side Channel was located in the lower portion of the side channel and was 1410 feet long (Appendix Figure D-36).Hydraulic information was collected from seven transects within this study site (Appendix Figures 0-37 &0-38).The channel within the study site has a gradual bend.The right bank from transects 2 to 6 is erosional,becoming less steep and depositional at transects 0 and 1.On the left bank,transects 2 through 6 are primarily deposi- tional in nature.In the areas of transects °and 1,the left bank becomes steep and erosional.At transect 2 on the left bank a small dewatered channel enters but water was never observed running in it (Appendix Figure 0-36).The thalweg gradient within the study site is 9.5 ft/mile (Quane et ale 1985).·Riparian vegetation along the right bank is primarily birch and spruce,whereas on the left bank it is alder. Transect °is located in a shallow pool habitat and has a substrate of sand and small gravel.Transects 1 (the discharge site)and 2 are primarily run habitat,and the substrate i~small gravel.At transect 3,the habitat changes to run and shallow pool habitat,the predominant substrate is small and large gravel.The hydraulic control for tran- sects 5 and 6 is transect 4.This transect represents riffle habitat, with substrates composed mostly of small and large gravels.Transects 5 and 6 are located in deep pool habitat,with small and large gravel substrate. D-64 - - - - ""'" - - ))})i l ))J ])}1 1 l Staff Gage Tr8.nsect €a River Mile o 1000 !I FEET (Approximate Scale) ...-.1'"~\'lE.0 SU 6 \1"'"d (BRM 87 o I 0'1 01 Appendix Figure 0-34.Overview of Sunset Side Channel (RM 86.4). / "/; ::o ...J U.ll Cl Wa: :Jen<w ~ Q.41.1 c fs 0...10-11•1*~1.- PZ'.O.."::o...J lL. Cl W 0: :J f/) <W ~ 0.<•10-o.sns rllSEL _'HIl '.]09' .2 •0.99 Q=45.8 chi----- / I /'""" - - -. SUNSET SIDE CHANNEL TRJ GAGE 86.951 n.lll1I)S Q i 101.000 <f, lfS[l •10-1•n66 Qo.n]3 •90 pI •O~" lOT C1IIITtlII.UO .11.11lO ~0 So J2.000 <'I VSEL •lOO....,J ao.ors,..90 pl.0".57 J--__CIlllTIOQI.lEO ...J Wen:: -II .._-;,+------.....---..1.-------_.- MAINSTEM DISCHARGE,SUNSHINE (x1000 cfs) •t -51gl ~f 01 Cl J'0 WSEL=93.30 .a 0> ~ -L;:!:ID _ MAINSTEM DtSCHARGE,SUNSHINE (x 1000 cfs) Appendix Fugure 0-35.Comparison of rating curves from Sunset Side Channel at transect 1 (from Quane eta al.1985). 0-66 Staff Gage Transect FEET (Approximate Scale) "",~".....o I ..J W,Z. "!Z-~;<::t: i 0 w Q f ,(; w (I)-a:z ~-(I) - Appendix Figure 0-36.Location of Sunset Side Channel study site (RM 86.9). 0-67 110 110 1011 CROSS SECTION 0 1011 JCROSS SECTION I 1011 STATION 0+00 1011 STATION 2+23...~ •107 ...107..•::lOll •1011 10' ... 10'-Z 104 1040Z 103 2 103 I-102«102 I- >101 «101 \IJ 100 >100.J \IJ UJ 1111 .J 1111 1111 \IJ 1111UJ117117>\IJ 1111 >1111I-115 --491.1.i=115« ..J 114 127 .t•«114 l&J 113 ..J 1130::\IJ 112 0::112 111 111 110 110 0 100 200 300 400 0 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEfT BANK HEADPIN (feet) 0 I 110 1100'1 1011 CROSS SECTION 2 1011 ~CROSS SECTION 3CO ~loa STATION 4 +75 loa STATION 7+58~...107 ...107•••1011 4)1011....... ~105 -105 Z 104 Z 104 0 103 0 103 I-102 i=102«101 «101>>\IJ 100 \IJ 100 .J 1111 .J 1111\IJ IIIl \IJ 1111 UJ 117 l&J 117>1111 >1111 ~V\;;::--i=4111.tI 5 ......-::::I 498 eh «115 127.h 115 --=127 el. .J 114-.J 114 \IJ 113 \IJ 1130::112 0::112 111 111 110 110 0 100 200 300 400 0 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN «feet)DISTANCE FROM LEFT BANK HEADPIN (feet) Appendix Figure 0-37.Cross section of transects 0,1,2,and 3 at Sunset Side Channel (adapted from Quane et al.1985).,J ..._J ),I ,J/J I J -] ~D J 'D J B J J ))1 }1 )-])l ")1 )J )1 -J 110 110 1011 CROSS SECTION 4 10111 CROSS SECTION 5 1015 STATION 9"'10 1011 STATION II ~53 ~ -;107 -107 1015 lD 1015..lD...105 -105~-104 1 Z 104 Z 10301030 5 102 ...102 101 ~et 101 >100 >100lIJW ..J 1111 ..J 99 lIJ 915 W 915 iii 97 W 87 ~,J 498 eft>911 498 eto >88 ---........127 eft--121 eft , i=95 ~~i=81S et 94 et 84..J ..J lIJ 93 W 113 0:112 0::92 111 91 110 110 a 100 200 300 400 a 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feetl DISTANCE FROM LEFT BANK HEADPIN (feet) 0 I 110en1011 ~CROSS SECTION 61..0 1011 STATION 14 ...10--101•J!1011....10e Z 104 2 103...102 ~101 W 100 ..J 911 lIJ 911 lIJ 117 ~, J 498 ell>811 , 127 ell"i=95et94..J lIJ 113 0::lI2 91 90 °100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feet) Appendix Figure D-38.Cross section of transects 4,5,and 6 at Sunset Side Channel (adapted from Quane et al.1985). Calibration Hydraulic data were collected at two calibration flows:127 and 496 cfs (Appendix Table 0-4).Mean daily discharges for the Susitna River on the dates that calibration data were collected at the Sunset Site Channel study site were 42,500 and 57,800 cfs,respectively.Based on these two calibration flows,an IFG-4 model was used to forecast instream hydraulics at this study site.The streambed profile,stage of zero flow,and observed and predicted water surface elevations for the study reach are plotted to scale in Appendix Figure 0-39.Both cali- bration data sets were used to predict hydraulic information from side channel flows of 7 to 1,603 cfs (mainstem discharges of 21,000 to 75,000 cfs)• To evaluate the performance of the hydraulic model,observed and pre- dicted water surface elevations,discharges,and velocity adjustment factors were compared (Appendix Table 0-11).The hydraulic model at Sunset Side Channel is similar to Circular Side Channel.Because of the 2 calibration flows,only a 2 point rating curve was formulated.In evaluating the performance of the model,observed and predicted WSEL1s and discharges were the same because of this rating curve.Velocity adjustment factors were all within the good range of 0.9 to 1.1. Additionally,the stage information of the model was compared to the rating curves established by Quane et ale (1985)(Appendix Figure 0-35). In the model,the stages of zero flow are not the same as those deter- mined from the thalweg survey by Quane et al.1985 (Appendix Table 0-12).The stage of zero flow values,input into the model,were derived from the thalweg points of the model input cross sections of transects 0,1,2,and 4.The reason for this change in thalweg eleva- tions is likely the result of the flood event.All the points used in the model were from measurements made,before the flood,whereas the Quane et ale (1985)thalweg survey was done after the flood -event. At transect 6,the velocities at the high calibration flow measurement (496 cfs)were adjusted upwards by 15%and at the low calibration flow measurement (127 cfs)adjusted downwards by 21%.Because this transect bisects a deep pool with eddies,it is difficult to obtain an accurate discharge measurement.The eddy effect was much more pronounced at the high calibration flow measurement,as there was about a 40 foot a section in which the velocities were negative.Because of its depth and slow velocities,this area was considered valuable habitat for rearing juvenile salmon.In order to facilitate using these negative velocity values in the model these measurements were treated as positive. At transect 3,there was a difference i nWSEL I s at the 127 cfs cali- bration flow.WSEL at the left bank was 95.03 feet,whereas at the right bank it was 94.90 feet.As the staff gage WSEL was 94.93 feet and the majority of flow occurred along this right side,a WSEL of 94.93 feet was used in the model. At transect 4,there was a large discrepancy (0.54 ft)in WSEL's across the transect at the .calibration flow of 127 cfs.This occurred because 0-70 '""" P'" - If/i!!!+, - 1 1 1 '))»)1 j 'J J 1 )))1 SUNSET SIDE CHANNEL Thalwl~Proflll with Ob.lrVldand Predictld Water Surfaci Profl/.. • ~ Thai."Surv.,Del.'840828 ThaI ••,Gradi.nl'8.11 f••lImil. Obatrv.d Waler Surfac.EI.varlan Slmu'at.d Wal.r Surfac.EI.vallan Eotrapalal.d Waf.r Surfac.EI.vallan Eleval ion af Z.ra I'la. Tho'••,Profile •• TII&"'ICT •n .....fel tn......cT 4ffll....UGT •nANueT itltAttllC!ITU"KCTO :_'_'_'_'_'_':'_'_'_'_'_'_'~:'_'_'_'_'_'_'';''::_'_'_'_'_._;~._:._._._...,._._.-1000 "I}_---=::::;.: ",.". .-_.-._.-.-_.__._._.----_._-_._._._._._-_.,..~~~.--.,~~~~_.-._._......-.-.,... _..._.~~~~wl!@~~~:~_~--~------~-."7JZ""W I'_I_~--W"~ •• .. .. !..o ~.. ..J.... >~ ..J.. III CJ I -....J..... ,I..1410 '••1-.1 n ;i ;iii Iii iii i'iii iii.iii iii Ii'iii ,iii i I Iii i ,iii iii'i ,iii Iii 001-00 noD IMOO 'StOO 10+00 U ....OO ]Otoo STREAMBED STATION 11001) Appendix Figure 0-39.Comparison of observed and predicted water surface profiles from calibrated model and surveyed thalweg profile at Sunset Side Channel (adapted from Quane et al. 1985). Appendix Table 0-11.Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Sunset Side Channel hydraulic model. D-72 - - - Appendix Table 0-12.Differences between stages of zero flow input into the model and Quane et al.(1985)thalweg survey at Sunset Side Channel. Stage of Zero Flow (ft) Transect Model Input Thalweg Survey 0 92.30 92.50 1 92.60 93.00 2 93.40 93.60 3 93.40 93.60 4 94.20 94.40 5 94.20 94.40 """6 94.20 94.40 ,.".. 0-73 the section of the channel where a majority of the flow occurred was higher in elevation and separated by a gravel berm from a lower eleva- tion minor channel where the staff gage was located.In order to utilize this cross section in the model,the channel cross section of the minor channel was elevated upwards by 0.6 feet. At a section of transect 3,because of channel configuration,the individual velocity measurements for the 127 cfs site flow were greater than the corresponding velocity measurements at the higher 496 cfs site flow.If these original values were to be used in the model the simu- lated velocities would decrease·with increasing site flows rather than increase as expected under normal circumstances.In order to amend this situation,the velocities were adjusted such that the relationship would simulate a positive increase in velocities with corresponding increases ins i te flow. Verification Based on the first level of verification by EWT&A,the model does an excellent job of simulating channel hydraulics between 50,000 and 61,000 cfs,mainstem discharge(275 and 649 cfs site flow).Above 61,000 cfs, the realiability of the simulated depth and velocity distributions begin to decrease.The model simulations were rated good between 61,000 and 64,500 cfs (649 and 850 cfs site flow),acceptable between 64,500 and 67,000 cfs (850 and 1,000 cfs site flow),and unacceptable above 67,000 cfs mainstem discharge.Below 50,000 cfs,the model simulations were also rated less than excellent,primarily because of reduced effec- ti veness in predi cti ng water surface profi 1es as compared to fi el d observations.The model simulations were rated good between 38,000 and 50,000 cfs (89 and 275 cfs site flow),acceptable between 32,000 and 38,000 cfs (41 and 89 cfs site flow),and unacceptable below 32,000 cfs mainstem discharge (Appendix Figure D-40). At the second level of verification there is excellent agreement for ve loci ty and good agreement for depth between observed and predi cted values (Appendix Figure 0-41).For a small number of depths there is a deviation away from the expected one to one relationship and this maybe attributable to the adjustments in the channel cross section at transect 4.The statistical tests show good agreement between these predicted and observed values (Appendix Table 0-7).The index of agreement is almost one,the total RMSE is mostly composed of the unsystematic RMSE, and the y-intercept is essentially zero with a slope of 0.99. Application For habitat simulation modelling purposes the hydraulic simulation model developed for Sunset Side Channel can simulate channel flows in the mainstem discharge range of 32,000 to 67,000 cfs. 0-74 - - - -1 1 )1 J J -]1 ].j 1 1 Application Range of the Calibrated Hydraulic Model at Sunset Side Channel RM (86.9) Si te Specific Flow.cfs o 7 31 107 280 614 1191 1603 I I I I I I ~~ ~ ~. ~~ ~~~~ I I I I I o 10 20 30 40 50 60 70 75 Cl I -.....J 01 Mainstem Discharge at Sunshine Station,cfs x 1000 _Excellent _Acceptable ill GoodoUnacceptable Appendix Figure 0-40.Application range of calibrated hydraulic model at Sunset Side Channel. SUNSET SIDE CHANNEL Observed V.s.Predicted Depths 4 r lffll 4 ata°3.5 -I Df(§I 3.5 D~~D ..-. 3 o~,0 0 3w,....III t ~mD "- ...."t2.5 'oDD'D '-'2.5I I-~lL ~DDW 0029 2 0 wwIfPdl0>I- 0 051.5 w 1.5wl- n:U lL 0 0 1 0 W I 0:: .........Il. O"l 0.5 D 0.5 o·0 0 1 2 :;4 0 OBSERVED DEPTH (FT) SUNSET SIDE CHANNEL Observed v.s.Predicted Velocities D ~ 2 :; 08SE~ED VELOCITY (fT /SEC) o #o 4 Appendix Figure D-41.Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 hydraulic mode1 at Sunset Side Channel. J J .~J J !!J J .~}J J 1 ,.". ~. I, Trapper Creek Side Channel (RM 91.6) Site Description Trapper Creek Side Channel is located on the west bank of the Susitna River and is approximately 5.0 miles long (Appendix Figure 0-42).It has a relatively uniform,broad,·and flat bottomed alluvial channel which is fed by multiple heads.It is separated from the mainstem Susitna River by a complex of sand bars,small channels,and vegetated islands.The head portion of this side channel is located in a complex of small channels and vegetated islands making it difficult to identify the origin of breaching flows (Quane et al.1985). During unbreached conditions,flows in Trapper Creek Side Channel are principally due to Cache Creek and groundwater from the upper reaches of the side channel.Breaching of Trapper Creek Side Channel is the result of the direct overtopping of the multiple heads of the side channel by the mainstem Susitna River.Based on assessments by Quane et al. (1985),the channel is estimated to be initially breached at a mainstem discharge of 43,000 cfs.Based ·on the comparison of the stage versus mainstem discharge rating curve for transect 4 (Appendix Figure D-43)by Quane et al.1985,a discharge of 44,000 cfs was selected as the con- trolling breaching discharge.This mainstem discharge corresponds to a streamflow measurement of 31.4 cfs. Based on assessments by Quane et a 1.(1985),backwater has not been observed.But at mainstem discharges ranging from 15,700 to 22,700 cfs, pooling was observed at transects 1,2,and 3 whi ch resul ted from the control located about 370 feet downstream from transect 1. The 790 foot long IFG modelling site at Trapper Creek Side Channel was located "in the lower portion of the side channel in a broad open channel .area (Appendix Figure 0-44).Four cross sections were surveyed within this area to define channel geometry (Appendix Figure D-45).The upper two transects were situated ina run,whereas the lower two transects were in a pool influenced by a downstream control.Substrate consisted primarily of rubble and gravels with some sand at the first transect. The thalweg gradient of the side channel is 12.1 ft/mile (Quane et a1. 1985). Calibration Hydraulic data were collected at three calibration flows:16,32,and 389 cfs (Appendix Table 0-4).Mean daily discharges for the Susitna River on the dates that cal ibration data were collected at the Trapper Creek study site were 20,900 cfs,44,000 ·cfs,and 57,700 cfs respec- tively.Based on these calibration flows an IFG-4 model was used to forecast instream hydraul ics for this study site.The streambed pro- file,stages of zero flow,and observed and predicted water surface elevations for the study reach are plotted to scale in Appendix Figure 0-46.All three data sets were used to predict hydraulic information for side channel flows from 9 to 1,351 cfs (mainstem discharges of 12,000 to 75,000 cfs). 0-77 Appendix Figure D-42.Overview of Trapper Creek Side Channel (RM 91.6). 0-78 - - - ;I'JSWJ',, - o Icoa ']'--]1 ]j J 1 I 1 j 1 1 ,;!....:il,;~,",;,@W"Y" ....~~ b.Staff Gage ~ ___Transect o 260 I 1 FEET (Approximate Scale) ]g Appendix Figure 0-44.Location of Trapper Creek Side Channel study site (RM 91.6). ! - _0 •10.3•5525 (lISEl •9DJ 11.6511 I Ie.%•D." 10 I~ tilI...... I~ I~ Q:31.4 ctl 0se •\D·%·D5l1 [VSD..1ll1 I.Z033--...... ,.-2 ..0." 3:o ...I LL.J Cl UJa:: :::l!J).«w::a: '.c •1O-IS •ts,\0..3."541 ,%•D."; r-__q,..IO..Z•SIlI3 0..'."]1 ,z •0." 3:o ....I LL.! o Wcr :::len« UJ::a: TRAPPER CREEK SIC TR4 GAGE 91.651 -.~;lfC ..,+--..J........_-.. MAINSTEM DISCHARGE.SUNSHINE (x1000 cfs)WSEL(+90 feet)' .t ~~:!o~r.,. ~I '-COITIOI.UD 15 0 I..l!6~£Ql oS o.s:l).lOD ell:: 2 ,o d I/S(t •IG"'''SS al •I 5&5 •to ~I ,Z·u '"'"'"WSEL=92.10 ....I ~"-.00.tD"'O.UID u-.ooc i CO So ....000 cfs wsn •IOD.DIllI OD.D'Y,•90 ...'..O~&J ~, - -..t-------~-------___. MAINSTEM DISCHARGE:SUNSHINE (x 1000 CfS) Appendix Figure D-43.Comparison of rating curves from Trapper Creek Side Channel transect 4 (from Quane et.al. 1985)• 0-79 j 1 I 1 )1 J J ·~l C~l i 10'105 104 CROSS SECTION I 104 j CROSS SECTION 2 103 STATION 0 ..00 103 STATION 2 ...89-102 102..ii..101:!:101 .. 100 ...100.... Z 88 Z 811081101111 I-117 j:117~>8S :!1111 UJ 115 II'.J UJ UJ 114 .J 9" 113 UJ 113~"-I 389 010 l&J 112 l&J 112 ...........<7 32 cro l6 01.>5 i=81 >III ~110 t-110 .J 118 ~118 l&J .J 11110::liS UJ 111 ct:111 1111 1111 S5 115 0 100 200 300 400 0 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet) 0 :~=j CROSS SECTION 4tlOll CROSS SECTION 3CO104.......103 STATION 5 +76 103 STATION 7+90--102 ...102-....101 ..101........100 ....100.... 811 Z 1111 Z III 0 8110117j:117~18 ~811 >85 UJ 85 l&J 84 .J 84.J UJUJ13 113,,~ ;;Ie 'P.'""'"160fa l&J 12 I&J 112 >81 >81 ~110 ~80 ..J 1111 .J elll l&J 1111 I&J 1111 D:117 II:117 III III I'115 0 100 200 300 400 0 100 200 300 400 DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (fset) Appendix Figure 0-45.Cross secti on of transects 1,2,3,and 4 at Trapper Creek Si de Channel . (adapted from Quane et al.1985). TRAPPER CREEK SIDE CHANNEL Thalwev Profile with ObServed and Predicted Water Surface Protlles .. /'L..!t0,.,1 IA"pro•.lco'" ~ Thalw.g Su,v.,Dol.'8401t11l Thalw.g G,adi.nt.12,1 f••I"nll. Ob ••rwed Wa'.r Surfao.Elevation SIIIlulaled Wat.,surfa..EI.vallan Elt,apalal.d Wal.,Surfa••EI.vaiion EI.valla.of Z.ra flail Thalw.g Profll. .. LI100'.0 GAGtl I'tISI IAO"'••O~Gf:"'51 AOF.'GAG£"DraB G.ol£ 1t.lIsl 'I"~ TIU.",tCT I TUNIECT J:TR'NSECT'TaANSlECl4 1011 "e'HI -It~!oo I I iii i a::t:oo i'i ,.iii 10';'00 ,i iii i I fe:':'OO ' .. =-.._._._._._._._._._._._.-._.-...O.,.}_."_._D_._._"_"_·---·-"•319 etl E.llapolaliop r01'l1l1 III'Z ......_._._._._._._._.._._.._._.._.._.••'52:c:f.",draulu:mod.' g .._._.._._....tach ~!c ta .:,.._._._._._._._._._._.._._._.._....._._.._._._.._.._._._._._._._._._.._....-.-._.-"'~".........-----*"'- ~.---------~------.----------------_.-------_._-~---------------------------_.-~~.J :10 ~ >~~ .J .... '" o I 00 N STREAMBED STATION Uootl Appendix Figure 0-46.Comparison of observed and predicted water surface profiles from calibrated model and surveyed thalweg profile for Trapper Creek Side Channel (adapted from Quane et al.1985). )-~I J __~t t .1 I ___.J --_.]_~_J - - i ,~ To evaluate the performance of the hydraul i c model,observed and pre- dicted water surface elevations,discharges,and velocity adjustment factors were compared (Appendix Table D-13).Of the 12 sets of observed and predicted WSEL1s,six sets were within ±0.02 feet of each other and the other six sets were within ±0.05 feet of each other.All the observed and predicted discharges were within 10%of each other except for'one set in which there was an 11%difference.All velocity adjustment factors were within the good range of 0.9 to 1.1.Addi- tionally,the stage information of the model was compared to the rating curves established by Quane et a1.(1985)(Appendix Figure D-43). Between the time that the first two calibration flows (389 and 32 cfs) were made and the last calibration flow of 16 cfs was made,the channel cross section at transect 1 was scoured by a flood event.In order to utilize this information in the model,the cross section determined from the survey and the 16 cfs flow measurement were used,and the WSEL I s of the two calibration flows (389 and 32 cfs)were then reduced by 0.37 feet. Transect 1 was determined to be a poor site for measuring discharge because it was a pool area affected·by a downstream control.The velocities for the 32 cfs calibration flow were therefore adjusted upwards by 27%,and at the 16 cfs calibration flow were also adjusted upwa rds by 20%. Verification Based on the first level of verification by EWT&A the model does a good job of simulating channel hydrau1 ics between 20,000 cfs and 54,000 cfs mainstem discharge (15 and 220 cfs site flow)(Appendix Figure 0-47). There are sufficient deviations in water surface elevation and discharge between predicted and observed values within this range to preclude attainment of the excellent rating.This occurs because the model is approximating a portion of the rating curve described by two adjoining linear relationships with a single line. Between 54,000 cfs and 58,000,cfs mainstem (220 and 460 cfs site flow) the model does an excellent job of simulating channel hydraulics. Beyond 58,000 cfs mainstem,the quality of the simulations begins to deteriorate as tne slope of the stage/discharge relationship for the site flattens with a change in channel geometry.The deviation between the regression l"ine developed within the model and that of the rating curve increases with discharge until the model simulations are no longer acceptable.The model simulations were rated good between 58,000 cfs and 61,000 cfs (460 and 600 cfs site flow),acceptable between 61,000 cfs and 66,000 cfs (600 and 820 cfs site flow),and unacceptable above 66,000 cfs mainstem (Appendix Figure 0-47). At the second level of verification there is good agreement between the observed and predicted values for velocity and depth (Appendix Figure 0-48).The statistical tests al so show good agreement between the predicted and observed values (Appendix Table 0-7).The index of agreement is 0.99,the total RMSE is largely composed of the unsys- tematic RMSE,and the y-intercept is almost zero with a slope near one. 0-83 Appendix Table D-13.Comparison between observed and predicted water surface elevations,discharges,and velocities for 1984 Trapper Creek Side Channel hydraul ic RI] model. - 0-84 J )1 J )'I 1 ]---]----1 1 Application Ranoe of the Cali brated Hydraul ic Mode I at Trapper Creek Side Channel RM (91.6) Site Specific Flow,cfs 1&22 29 101 564 1030 1351 I I I I I I 11:::== I I I I I I , o 10 20 30 40 50 60 10 15 CJ I OJ U1 Mainstem Discharoe at Sunshine Station.cfs x 1000 III Excellent _Acceptable •Good D Unacceptable Appendix Figure 0-47.Application range of the calibrated hydraulic model at Trapper Creek Side Channel. TRAPPER CREEK SIDE CHANNEL TRAPPER CREEK SIDE CHANNEL Observed vs.Predicted Depths 4 ,~ o~jo,0 o ~08 2 .3 OBSERVED VELOCITY (FT!SEC) Observed vs.Predicted Velocities 4 I I 3.5 ,.... 0 .3w lfl "-t 2.5'-' ~ £3 20 ..J ~ 0 1.5wt-o iswn: 0- 0.5 0 02.82.4 o 8 o 0 0008 0 IJD o~Q DtB~D °fQ1 Co o 0.8 1.2 1.6 2.0 OBSERVED DEPTH (fEET) 0.4 3 i I 2.8 2.6 2.4 2.2 2 1.B 1.6 1.4 1.2 0.8 0.6 0.4 0.2 o 0.0 ~w W IL. '-' :r twa owt- U iswn: 0.o I OJ 0'1 Appendix Figure D-48.Scatterplots of observed and predicted depths and velocities from the calibrated IFG-4 model at Trapper Creek Side Channel. .)J ~.J _J .J ,~_....,J _c.~)__I Application For habitat simulation modelling purposes the hydraulic simulation model developed for Trapper Creek Side Channel can simulate channel flows in the mainstem discharge.range of 20,000 to 66,000 cfs. SUMMARY A summary of the range of Illainstem discharges that the hydraulic models can simulate for the rearing habitats of salmon at the six lower river IFG modelling sites is presented in Appendix Table 0':'14. I~ Appendix Table 0-14. Site (RM) Summarization of the range of mainstem discharges that the hydraulic models can simulate for the rearing habitats of salmon at the six lower river IFG modelling sites. Mainstem Discharge Range (cfs) - ,.... Island Side Channel (63.2) Mainstem West Bank (74.4) Circular Side Channel (75.3) Sauna Side Channel (79.8) Sunset Side Channel (86.9) Trapper Creek Side Channel (91.6) 0-87 35,000 to 70,000 18,000 to 48,000 36,000 to 63,000 44,000 to 63,000 32,000 to 67,000 20,000 to 66,000 ..- ACKNOWLEDGEMENTS The authors express their appreciation to the following for their assistance in preparing this report: The other ADF&G Su Hydro Aquatic Studies Program staff who provided their support to this study.For collection of field data:Fred Metzler,Pat Morrow,Isaac Queral,Glenn Freeman, and John McConnaughey.To Paul Suchanek for collection of the cover information used in assessing the weighted usable areas of the models.In reduction of the data,making the many compu~er runs,and helping prepare this appendix:Fred Metzler,Mary Shiffer,Dan Kingsley,and Kathy Sheehan Dugan. To Tim Quane,Pat Morrow,and Isaac Queral for use of much of their findings and figures from Task 36 support technical report -Hydrological Investigations at Selected Lower Susitna River Study Sites.To the editors:Doug Vincent-Lang,Tim Quane,and Drew Crawford.For the cartography on the final figures:Carol R.Hepler and Roxann Peterson. To E.W.Trihey and Associates;particularly Bob Aaserude and Diane Hilliard for their valuable expertise in the collection of data,the calibration,and the verification of the hydraulic models . 0-88 ..... 1, j ..... - I..... LITERATURE CITED Acres American,Inc.(Acres).1982.Susitna hydroelectric project draft FERC license application,volume 1,exhibit E,chapter 2. Anchorage,Alaska • Ashton,W.S.,and S.A.Klinger-Kingsley.1985.Response of aquatic habitat surface areas to mainstem discharges in the Yentna River confluence to Talkeetna reach of the Susitna River.Draft report. R&M Consultants,Inc.and E.Woody Trihey and Associates.Prepared for Alaska Power Authority.Susitna Hydroelectric Power Project. Anchorage,Alaska. Bovee,K.D.,and R.Milhous.1978.Hydraulic simulation in instream flow studies:theory and techniques.Instream Flow Information Paper No.5.Instream Flow Service Group.USFWS.Ft.Collins, Colorado. •1982.A guide to stream habitat and analysis using instream-----~f'low incremental methodology.Instream Flow Information paper No. 12.Coop.Instream Flow Service Group.USFWS.Colorado. Buchanan,T.J.,and W.P.Somers.1969.Techniques of water resources investigations of the United States Geological Survey.Chapter A8. Discharge measurements at gaging stations.USGS.Washington DC. Hilliard,N.D.1985.Extrapolation limits of the 1984 middle river IFG models.Technical Memorandum.E.Woody Trihey and Associates. Anchorage,Alaska. Hilliard,N.D.,S.Williams,E.Woody Trihey,R.C.Wilkinson,and C.R.' Steward III.1985.Summary of hydraulic conditions and habitat forecasts at 1984 middle river study sites.Draft report.E. Woody Trihey and Associates.Prepared for Alaska Power Authority. Susitna Hydroelectric Power Project.Anchorage,Alaska. Instream Flow Group (IFG).1980.The incremental approach to the study of instream flows.USF&WS.W/IFG-8)W31.Ft.Collins,Colorado. Main,R.1978.IFG-4 program user's manual.U.S.Fish and Wildlife Service.45 pp. Milhous,R.T.,D.L.Wegner,and T.Waddle.1984.User 's guide to the Physical Habitat Simulation System (PHABSIM).U.S.Fish and Wildlife Service.Instream Flow Information Paper No.11. FWS/OBS-81/43 Revised.Fort Collins,Colorado. Quane,T.,P.Morrow,and I.Queral.1985.Hydrological Investigations at Selected Lower Susitna River Study Sites.Alaska Department of Fish and Game.Su Hydro Aquatic Studies Task 36 Support Technical Report.Alaska Department of Fish and Game.Anchorage,Alaska. D-89 Suchanek,P.M.,K.J.Kuntz,and J.P.McDonell.1985.The relative abundance,distribution,and instream flow relationships of juvenile salmon in the lower Susitna River.Alaska Department of ~ Fish and Game.Susitna Aquatic Studies Report No.7,part 2. Alaska Department of Fish and Game.Anchorage,Alaska. Trihey,E.W.1979.The IFG incremental methodology.In G.L.Smith, ed.Proceedings of the Instream Flow Criteria and Modeling Workshop.Colorado Water Resources Research Institute,Colorado State University.Pages 24-44.Information Series No.40.Fort Collins,Colorado. •1980.Field data reduction and coding procedures for use with --"7""t'he IFG-2 and IFG-4 hydraul ic simul ati on model s.Instream Flow Service Group,USFWS.Fort Collins,Colorado. •and D.L.Wegner.---use with the physical Flow Group.Instream Colorado. 1981.Fiel d data collection procedures for habitat simulation system of the Instream Flow Service Group.USFWS.Fort collins, Wilmott,C.J.1981.On the validation of models,physical geography 2. V.H.Winston and Sons.p.184-194.~ - ..... - D-90