£93423;

In: .

. ..z . V

315...!
o. 7'. L

. . imam.
"97.1.2. “I! .

a”:
,. a

.12.: ., .
n... . .

 

“ﬂaunt. .«nld.»
slith... q
.f‘E. .:.!....t.
.1. , v3.53;
Q... ea... .
Nan
iaﬂ. .
ﬁx. 9“
Izciﬁmhv 214-... .
ﬂintuuhrsﬁﬂuaﬁn
.z,....u.. “wax ‘ It
3.2.7.3 .

.9!
.1
23“, 3..- .. .4
.wmlusm

{.1

 

 

xhnwuila: T 1.: . 5.:
ﬁimﬁ _ ‘ ‘ .

.wiﬁi. .

 

THESS
j

‘-

”Illilllllllll‘nlll‘llllllli‘ll ”

31293 01690 3191

This is to certify that the

thesis entitled

Feeding Habitats and Condition of Juvenile Chinook
Salmon in the Upper Sacramento River, California

presented by

Pamela Ann Petrusso

has been accepted towards fulﬁllment
of the requirements for

M.S. Fish. & Wildl.

degree in

WWW

Major professor

 

 

Date H Gui/7f I???

07639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINE-S return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1198 Mamba-p.14

FEEDING HABITS AND CONDITION OF JUVENILE CHINOOK SALMON IN
THE UPPER SACRAMENTO RIVER, CALIFORNIA

By

Pamela Ann Petrusso

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Fisheries and Wildlife

1998

ABSTRACT

FEEDING HABITS AND CONDITION OF JUVENILE CHINOOK SALMON IN
THE UPPER SACRAMENTO RIVER, CALIFORNIA

By

Pamela Ann Petrusso

The feeding habits, available foods, and physical condition of juvenile Chinook
salmon Oncorhynchus tshauytscha rearing in the upper Sacramento River during 1995
and 1996 were examined. Numerically dominant taxa in driﬁ samples and stomach
contents included Chironomidae (Diptera) and Baetidae (Ephemeroptera). Juvenile
chinook selected baetid adults, chironomid adults, and chironomid pupae, and avoided
baetid nymphs, chironomid larvae, and other small insects and crustaceans, based on linear
selection indices. The gape width of juvenile salmon increased predictably with ﬁsh size,
allowing larger ﬁsh to potentially exploit larger maximum prey sizes; however, larger
salmon continued to include small items in their diet to maintain adequate energy intake.
Condition factor (K) increased and percent body water decreased with salmon size due to
allometric length-weight and length-dry weight relationships. Mean length, weight, and
condition factor increased, and mean percent water decreased, over time (P < 0.0001) and
from upstream to downstream sites (P < 0.0001 ). Analysis of covariance showed weight
at a given length increased progressively from February through June 1996 for salmon 50
to 90 mm. F ield-caught salmon in general were heavier at all lengths than salmon

experimentally fasted for two weeks.

ACKNOWLEDGEMENTS

I would like to thank my advisor, D. Hayes, and committee members, C. McNabb
and R Merritt, for their guidance. Thanks also to J. Smith at the US. Fish and Wildlife
Service, Northern Central Valley Fishery Resource Ofﬁce, Red Bluff, California, and L.
Hanson at the California Department of Fish and Game, Environmental Services Division
Ofﬁce, Red Bluﬂ‘, for allong me to accompany their crews during ﬁeld sampling. I am
grateful to the many individuals who provided valuable advice or help with permits, ﬁeld
work, or lab work, including: S. Borthwick, R. Crouch, D. Goebel, W. Hickman, L.
Holsinger, J. Lessard, T. Moore, T. Newcomb, K. Newman, K. Perry, J. Renaud, W.
Snider, M. Stodolski, T. Thompson, D. Traeumer, B. Weilmunster, and M. Weisenreder.
C. Liston and C. McNabb reviewed earlier versions of this document; their insights are
greatly appreciated.

Special thanks to my husband, W. Talo, whose commitment of time, energy,
ﬁnances, and emotional support was vital to the completion of this project.

Funding was provided by a grant from the US. Bureau of Reclamation, Ecological
Research and Investigations Group, Denver, Colorado, under cooperative agreement

number l425-4-FG—8 1-20040.

TABLE OF CONTENTS

 

 

 

 

LIST OF TABLES .................................................... vii
LIST OF FIGURES ................................................... x
INTRODUCTION ..................................................... 1
The Study Area .................................................. 3
Life History of Fall-run Juvenile Chinook Salmon ....................... 6
METHODS .......................................................... 8
Study Sites and Duration .......................................... 8

Drift Collection and Handling ..................................... 1 1

Field Collections of Invertebrate Drift .......................... 11

Processmg' of Drift Smles ................................. 12

Fish Collection Methods .......................................... 13

Beach Semi—ng' ' ............................................ 13

Other Methods of Fish Capture ............................... 14

Diet Analysis .................................................. l4

Drift and Feeding over the Diel Cycle .......................... 16

Condition Estimates ................ A ............................. 17
Condition Factor .......................................... 17

Du Weight and Percent Water ............................... l7

 

iv

Statistical Methods .............................................. 18

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

RESULTS .......................................................... 20
Invertebrate Drift and Salmon Diet ............................. '. . . . 2O
Drift Density ............................................. 20
mm Abuadances of Drift Taxa ............................ 20
Relative Abundaaces of Taxa in the Diet ........................ 22
Frequency of Occurrence in the Diet .......................... 24
Selection of Available Prey by Juvenile Chinook .................. 24
EisrhSize. Gape Size and Prey Size Rel_ationships .................. 27
Feed'mg Intenaityand Stomach Fullness ......................... 29

Drift ﬂd Feeding over the Diel Cycle .......................... 29
Condition of Juvenile Chinook Salmon ............................... 36
Length-Weight and Length-Common Relationships ............... 36
Length-Dry Weight Relationship .............................. 36
General Linear Models Analysis of Spatial and Temp_oral Patterns ..... 36
Aialysis of Covariance of Regressions by Month and River Mile ...... 37
DISCUSSION ....................................................... 47
Invertebrate Drift and Salmon Diet ................................. 47
Drift Densig ............................................. 47
wive Abundances of Tax_a in Drift anQDiet ................... 48
Frequency of Occurreng ................................... 49

Prey Selection ............................................ 49

 

V

Fig Size. Gape Size, and Prey Size Relationshipa ................. 50

 

 

 

 

 

 

 

 

 

Feeding Intensity and Stomach Fullness ......................... 51

Md Feeding over the Diel Cycle ...................... A. . . . 52

PM Sources of Error ................................... 53

Condition of Juvenile Chinook ..................................... 54
Length-Weight Relationship ................................. 54

Qagth-Dry Weigh; Relationship .............................. 56

Condition Factor (K) ....................................... 57

General Linear Models Analysis of Spatial and Temral Patterns ..... 59

Mysis of Covariance of Length-Weight Relationahipa ............ 59

Possible Sources of Error ................................... 60

SUMMARY AND CONCLUSIONS ...................................... 6l
APPENDICES ....................................................... 64

APPENDIX A. Relative Abundances of Drift Taxa July-August 1995 and
January-March 1996 ....................................... 64

APPENDIX B. Relative Abundances of Drift and Diet Taxa April-June 1996 . .70

APPENDIX C. Mean Length, Weight, Condition, and Percent Water of

Sacramento River Juvenile Chinook Salmon ..................... 76

APPENDIX D. Growth Experiments with Hatchery Salmon .............. 78
APPENDIX E. Experiments on Postmortem Changes in Length and

Weight ................................................. 86

LIST OF REFERENCES ............................................... 89

vi

LIST OF TABLES

Table 1. Life history characteristics of the four stocks of Sacramento River chinook
salmon. Information is compiled from Vogel and Marine (1991) and Fisher (1994) . . . . 7

Table 2. Sites used for drift and ﬁsh sampling on the upper Sacramento River, California,
1995-1996. Sites are listed in order ﬁom downstream (Bidwell Park) to upstream
(Caldwell Park) ....................................................... 10

Table 3. Mean, standard deviation, and 95% conﬁdence for drift density (number of
organisms per 100 in3 water ﬁltered per net) by site and date for the upper Sacramento
River, 1995-1996. RM = river mile. N = number of samples (nets) taken each date . . . 21

Table 4. Percentages of common taxa found in the drift and diet of juvenile chinook
salmon in the upper Sacramento River during 1995 and 1996 (n = number of drift or
stomach samples examined, n. = total number of items identiﬁed and used to estimate
abundances). Drift percentages recalculated without the “other” category are included in
parentheses (see text for explanation) ...................................... 23

Table 5. Frequency of occurrence (number of stomachs containing one or more items of a
taxon) of selected prey taxa of juvenile chinook salmon captured from the upper nminstem
Sacramento River, April-June 1996. Only items present in over 10% of stomachs are

listed .............................................................. 25

Table 6. Mean, standard deviation, maximum, and minimum head capsule widths of
common taxa found in stomachs of juvenile chinook salmon captured in the upper
Sacramento River, April-June 1996 ....................................... 29

Table 7. Mean number of prey per stomach and mean stomach fullness index (SFI) of
juvenile chinook salmon captured in the upper Sacramento River, April-June 1996,
arranged by month and size class, and presented with approximate 95% conﬁdence limits.
................................................................... 30

Table 8. Daytime and nighttime taxonomic composition of the Sacramento River drift
during one 24-hour period, 19-20 April 1996 ................................ 35

Table 9. Results of general linear models analyses, describing the effect of month and
river mile of capture on length, wet weight, condition factor (K) and percent body water
of juvenile chinook salmon captured from the Sacramento River, 1995-1996 (n = 1204 for
length, weight, and K; n = 520 for percent water) . . . .- ........................ 42

Table 10. Least squares mean length, weight, condition factor (K), and percent water
(presented with standard errors) of Sacramento River juvenile chinook by month/year of
capture, averaged across all sites. The estimates were generated by analyses with general
linear models. Length, weight, and condition data based on delayed measurement were
adjusted to reﬂect live measurements (Live length = delayed length/(0.9844); live weight =
delayed weight/(10287)). Percent water was calculated using the thawed weights of ﬁsh
rather than the original live weights. N.a. = not applicable because ﬁsh were not sacriﬁced
during those months ................................................... 43

Table 11. Least squares mean length, weight, condition factor (K), and percent water
(presented with standard errors) of Sacramento River juvenile chinook by site (river mile),
averaged across months. Estimates were generated by analyses with general linear
models. Length, weight, and condition factor data based on delayed measurement were
adjusted to reﬂect live measurement (Live length = delayed length/(0.9844); live weight =
delayed weight/(10287)). Percent water was calculated based on the thawed wet weight
and dry weight of juveniles captured April-June 1996. Na. = not applicable, sample not
taken .............................................................. 44

Table 12. Slopes and intercepts for In length-In weight regressions (for ﬁsh 50 to 90 mm
FL) and length-percent water regressions by month for juvenile chinook salmon captured
ﬁom the Sacramento River, 1995-1996. Pooled slope estimates were determined by slope
heterogeneity tests. Intercepts, determined by analysis of covariance, are ranked ﬁ'om
highest to lowest intercept for weight, and from lowest to highest intercept for percent
water ............................................................... 45

Table 13. Slopes and intercepts for In length-1n weight regressions (for ﬁsh 50 to 90 mm
FL) and length-percent water regressions by river mile for juvenile chinook salmon
captured from the Sacramento River in 1996. Pooled slope estimates were determined by
slope heterogeneity tests. Intercepts, determined by analysis of covariance, are ranked
ﬁ'om highest to lowest intercept for weight. Both length-water regression slopes and
intercepts are reported, as the slopes were found to be signiﬁcantly heterogeneous . . . . 46

Table 14. Total numbers and relative percentages of organisms captured in the drift from
the Sacramento River, July-August 1995 (n = 84 samples), and late January-March 1996
(n = 74 samples) (“-” indicates item not present) ............................. 64

Table 15. Relative abundance of items in the drift (n = 113 samples) and diet of juvenile
chinook (n = 189 ﬁsh) in the Sacramento River for the period April-June 1996. Two ﬁsh
had empty stomachs (“-” indicates an item not present) ........................ 7O

viii

Table 16. Mean fork length ( +/-SE), mean weight (+/- SE), mean condition factor (K)
(+/-SE), and mean percent water (+/—SE) (where applicable) of young-of-year chinook
salmon captured at study sites on the upper Sacramento River, July 1995 and F ebruary—
June 1996. Mean length, weight and condition reported for certain sites sampled in April
and May were adjusted for delays in measurement 4-11 hrs postmortem. In cases where
the number of ﬁsh used for determination of percent water is different from the number
used for length, weight and condition calculations, the n for percent water is reported in
parentheses. Mean percent water was calculated based on the thawed weight of ﬁsh taken

just prior to drying ................................................... 76

Table 17. Summary of treatment information for growth experiments with juvenile
chinook salmon ...................................................... 8O

LIST OF FIGURES

Figure 1. Map of the study area and location of study sites on the Sacramento River,
1995—1996. Inset: Map of California showing location of the city of Red Bluff. Redrawn
from the US. Bureau of the Census, Tiger Mapping Service Database, 1997 ........ 9

Figure 2. Linear selection indices (L) for selected taxa found in the drift and the diet of
juvenile chinook salmon in the upper Sacramento River, April-June 1996. Possible values
of L range from 1 to —1 ................................................ 26

Figure 3. Regressions of gape width, and maximum, mean, and minimum head capsule
width of invertebrate prey, on fork length of juvenile chinook salmon captured April-June
1996 ﬁom the upper Sacramento River ..................................... 28

Figure 4. Mean drift density (number of organisms per 100 m3) and stomach fullness
index (SF 1) (% of total ﬁsh wet weight) of juvenile chinook salmon captured from the
Sacramento River, 19-20 April 1996. Sampling began at 2000 hours PDT (hour 0). . . 32

Figure 5. Mean drift density (number of organisms per 100 m3) and mean number of prey
per stomach of juvenile chinook salmon captured from the Sacramento River over a 24-
hour period, 19-20 April 1996. Sampling began at 2000 hours PDT (hour 0) ........ 33

Figure 6. Frequency distribution of invertebrates of different size classes (head capsule
width) present in the drift and juvenile chinook salmon diet over a 24-hour period, 19-20
April 1996, in the Sacramento River ....................................... 34

Figure 7. Ln fork length (mm) vs. ln weight (mg) of juvenile chinook salmon captured
from the Sacramento River, California, 1995-1996 (11 = 1204). Lengths and weights were
adjusted for delayed measurements where appropriate .......................... 38

Figure 8. Relationship between condition factor (K) and fork length of juvenile chinook
salmon captured from the Sacramento River, 1995-1996 (n = 1204). Length and
condition were adjusted for delayed measurement where appropriate ............... 39

Figure 9. Ln length (mm) vs. ln weight (mg) of juvenile chinook salmon captured from
the Sacramento River, California, 1995-1996, compared with salmon ﬁ'om growth
experiments (Level 0 = unfed two weeks; levels 1 and 2, fed) .................... 4O

Figure 10. Relationship between In fork length (mm) and In wet weight (mg) or In dry
weight (mg) for juvenile chinook salmon captured April-June 1996 from the Sacramento
River (11 = 520) ....................................................... 41

Figure 11. Initial and ﬁnal (after two weeks) In length-1n weight relationships for the three
treatment groups in Growth Experiment 1. Level 0 = zero ration treatment; Levels 1 and
2 were fed. One outlier was excluded from level 0 ............................ 83

Figure 12. Initial and ﬁnal (after two weeks) In length-1n weight relationships for the two

treatment groups in Growth Experiment II (Level 0 = zero ration treatment; Level 1 was
fed) ................................................................ 84

xi

INTRODUCTION

Juvenile chinook salmon Oncorhynchus tshauytscha spend from several weeks to
several years rearing in freshwater before migrating to the ocean. For chinook in the
Sacramento River, most of which migrate as subyearling fry or smolts (“ocean-type”
salmon, Healey 1983), residence in upper river habitats can be short. However, growth in
size and stored energy during early rearing can ultimately determine survival to the
estuarine or ocean phase. Consideration of important features of the environment, such as
food availability, along with the response of the ﬁsh (e.g., food selection and overall ﬁsh
condition) to habitat features, allows for a greater understanding of the role of freshwater
rearing in the life history of juvenile chinook salmon.

Four chinook salmon stocks use the upper Sacramento River and its tributaries for
spawning: fall, late-fall, winter and spring runs. All four stocks have declined in an era of
numerous anthmpogenic changes to the river such as construction of dams, ﬂow
regulation and diversion, stream channelization, conversion of vast areas of ﬂoodplain to
agriculture, and gravel extraction, among others (Brown 1991, Fisher 1994). Winter-run
chinook salmon were recently listed as endangered under state and federal endangered

species laws (National Marine Fisheries Service 1994), the spring run is the focus of

current efforts for endangered species protection, and the late-fall and fall stocks rely

2
heavily upon hatchery supplementation. Despite the many research projects taking place

on the Sacramento River to increase understanding of salmon life history and production,
condition of juvenile chinook rearing in the upper river has received little attention
(Kjelson 1993). Recent studies focusing on food availability and feeding habits of
juveniles in the upper mainstem Sacramento River have also been lacking.

Information on the physical robustness, or condition, of juvenile chinook salmon
serves two important purposes. Condition indices provide an assessment of how well ﬁsh
are coping with their environment (Goede and Barton 1990). Condition indices therefore
reﬂect how suitable the environment is for rearing. When viewed as a general measure of
energy status or as an indicator of stress (Love 1970, Adams 1990), condition indices may
additionally provide insight into the competency of the ﬁsh to meet future challenges to
survival such as food shortages and other acute or chronic stressors. To the extent that
measures of ﬁsh condition provide a view of both the rearing history and possible future
response of a ﬁsh to various challenges, baseline information on condition can be of use to
ﬁshery managers in interpreting results of their studies on juvenile salmon in the
Sacramento River.

Pairing a description of the available foods and diet of juvenile chinook salmon
with information on ﬁsh condition-~toward the goal of determining general suitability of
the rearing habitat--has several advantages over studies focusing on ﬁsh feeding or ﬁsh
condition alone. Fish feeding represents a short-term (several hours) response to
environmental conditions, whereas ﬁsh condition reﬂects a relatively longer term,
integrated response (several days to weeks). Thus the two kinds of data complement each

other. The density, size, and taxonomic composition of available foods in relation to food

3
selection and feeding intensity all aﬂ°ect ﬁsh energy intake and expenditure. Energy

acquisition at least partially determines ﬁsh condition; however, extreme estimates of ﬁsh
condition cannot be attributed to food limitation or food excess without some evidence of
this from food and feeding information. Similarly, ﬁsh feeding habits studies that report
high food production and high feeding intensity cannot support an assertion of habitat
suitability for growth in size or energy without information on ﬁsh condition or growth
rates.

The main objectives of this study were to: (1) assess the seasonal availability of
natural foods (invertebrate drift) in upper mainstem habitats of the Sacramento River; (2)
describe the feeding habits of juvenile chinook salmon in the upper river; (3) determine
the physical condition of juvenile chinook salmon in upper river habitats; and (4)
determine patterns in condition of juvenile salmon over time, and from upstream to
downstream sites. Because of the low abundance of late-fall, winter, and spring-run
stocks, fall-run young-of-year were used for determinations of condition or feeding habits

which required sacriﬁce of juvenile salmon.

The Study Area
The Sacramento River originates near Mt. Shasta in northern California, and
travels along the ﬂoor of the Central Valley to its mouth in the Sacramento-San Joaquin
Delta for a total of approximately 370 river miles (RM). The northern Central Valley is
bordered by the Sierra Nevada and Cascade ranges on its east side and the Klamath and
Coast ranges on its west. The climate is characterized by heavy precipitation during

winter and spring of wet years, and little or no precipitation during summer and fall.

4
Fed primarily by runoff from precipitation and snowrnelt ﬁom surrounding

mountains, the Sacramento River at Red Bluﬂ historically exhibited ranges in ﬂow from
about 28 cubic meters per second (m3/s) in summer to 100 year ﬂood levels of about
11,893 m3/s (California Department of Water Resources 1994b). River discharge is now
regulated by the Shasta-Keswick Dam complex (RM 311 and RM 302), the operation of
which tends to dampen oscillation in the hydrograph, as compared to historic ﬂows
(California Department of Water Resources 1994a). River ﬂow at a given site largely
depends upon the management of water projects upstream. Major water diversions
located in the study area include the Anderson-Cottonwood Irrigation Diversion (ACID)
(RM 299), the Red Bluff Diversion Dam (RBDD) (RM 243), and the Glenn-Colusa
Irrigation Diversion (GCID) (RM 206). The river provides the main water supply for the
cities of Redding, Chico, and Red Bluff, in addition to supporting agricultural and
recreational water uses.

The study area, from RM 193 to RM 299, consists of two reaches of the river
characterized by distinct morphologies (Buer 1984). Upstream of Red Bluﬂ‘, the river is
primarily a bedrock stream controlled by the underlying geology. Downstream of Red
Bluﬂ‘, the river morphology is alluvial, primarily controlled by discharge and sediment
load.

The riparian habitats of the Sacramento River can be classiﬁed into four lateral
zones (California Department of Water Resources 1994a). Gravel bars are characterized
by sparse stands of young willow (Salix spp.) and various forbs and grasses. Riparian
scrub, which vegetates the area from the gravel bar to the high-water edge of the channel,

also is comprised of young woody vegetation as well as brush and debris piles. Riparian

5

forests in ﬂoodplain and terrace areas outside the active channel bed consist of understory
vegetation such as poison oak Rhus diversiloba, blue elderberry Sambucus mexicana,
California blackberry Rubus vitifolius, willow Salix spp., wild grape Vitis californica, and
forbs and grasses. The overstory vegetation is comprised primarily of cottonwood
Populusfremontii, California sycamore Platanus racemosa, box elder Acer negundo,
Sierra alder Alnus rhombifolia, black walnut Juglans hinsii, valley oak Quercus lobata,
and interior live oak Quercus wislizenii (California Department of Water Resources
1994a). The riparian forest has been reduced to 5% of its historic land area of 500,000
acres, replaced by irrigated agriculture (California Department of Water Resources
1994b). The fourth riparian habitat includes the artiﬁcial and natural ponds, oxbow lakes,
and backwaters which provide wetland habitat on an ephemeral basis.

The ﬁsh fauna of the Sacramento River includes a number of native and introduced
species. Hardhead Mylopharodon conocephalus, Sacramento perch Archoplites
interruptus, tule perch Hysterocarpus traski, Sacramento sucker C atostomus occidental is,
Sacramento squawﬁsh Ptychocheilus grandis, splittail Pogonichthyes macrolepidotus,
hitch Lavina exilicauda, Sacramento blackﬁsh Orthodon microlepidotus, California roach
Hesperoleucus symmetricus, Paciﬁc lamprey Lampetra tridentata, white sturgeon
Acipenser transmontanus, steelhead/rainbow trout Oncorhynchus mykiss, speckled dace
Rhinichthys osculus, prickly sculpin Cottus asper, and riﬂle sculpin Cottus gulosus are
native inhabitants of the river (Moyle 1976). Striped bass Morone saxatilis and American

shad Alosa sapidissima are examples of introduced species with in-river sport ﬁsheries.

6
Life History of Fall-run Juvenile Chinook Salmon

F all-run chinook in the study area emerge ﬁom redds during December through
March of each year, with peak emergence at the end of January (Vogel and Marine 1991)
(Table 1). After emergence, the ﬂy seek habitats with slow currents and adequate cover
and begin to feed upon insects in the drift (Moyle 1976). During years characterized by
heavy precipitation and associated high ﬂows and high turbidity, fall-run ﬁy emigrate from
the upper river system almost immediately after emergence (Vogel and Marine 1991).
During dry years, early emigration is low, and most fry remain to rear in the upper river
through mid-April or mid-May. Though most fall-run juveniles are thought to exit the
upper river by the end of June, a few 90 to 200 mm salmon are captured in the
Sacramento-San Joaquin Delta during the fall. These catches probably include members
of the fall-run that spent the summer in upper river habitats (Kjelson et al. 1982).

F all-run ﬁ'y emigrating from the upper river reach peak abundance in the
Sacramento-San Joaquin Delta in February or March of wet years (Kjelson et al. 1982).
Although the importance of upper-river versus Delta-estuarine rearing is still under debate,
there is evidence that salmon rearing in the Delta may grow more quickly than juveniles
rearing in the upper river during the same period (Kjelson et a1. 1982). Upper-river
rearing for the fall-run may be more critical during dry years, when outmigration of

juveniles is delayed.

7

Table 1. Life history characteristics of the four stocks of Sacramento River chinook
salmon. Information is compiled from Vogel and Marine (1991) and Fisher (1994).

 

 

Characteristic Fall Run Late Fall Run Winter Run Spring Run
Adult Migration July-Dec Oct-April Dec-July March-Oct
Peak Sept-Oct Dec March May-J une
Spawning Oct-Dec Jan-April April-Aug Aug-Oct
Peak Late Oct Early Feb Early June Mid-Sept
Incubation and Oct-Mar Jan-June April-Oct Aug-Dec
Emergence
Juvenile rearing Dec-June April-Dec July-Mar Nov-May
and outmigration
Freshwater 4-7 months 7-13 months 5-10 months 3-15 months
residency
Ocean entry of Mar-July Oct-May Nov-May March-June and
juveniles Nov-March
Size at ocean 80 mm FL 160 mm FL 120 mm FL 80 mm FL
entry

 

Hatchery-reared fall, late-fall, and winter chinook salmon are released during

winter and early spring from Coleman National Fish Hatchery, located on Battle Creek, a

tributary of the Sacramento River. During the study period in early 1996, approximately

7.5 million unmarked fall-run fry, and 12.3 million fall-run smolts (about 8% marked with

adipose ﬁn clips and coded wire tags) were released into Battle Creek or into the

mainstem Sacramento River downstream of the Red Bluﬂ‘ Diversion Dam at RM 242 (S.

Croci, US. Fish and Wildlife Service, Northern Central Valley Fishery Resource Oﬂice,

Red Bluﬂ', California, personal communication).

METHODS

Study Sites and Duration

Investigations of the available foods, feeding habits, and condition of juvenile
chinook salmon were conducted on the upper mainstem Sacramento River between the
cities of Redding (RM 299) and Chico (RM 193), California (Figure 1) during July-
August 1995 and late January-June 1996. Juvenile fall-run chinook salmon were
sacriﬁced for estimates of dry weight and feeding habits from April-June 1996 only.

A complete list of study sites is given in Table 2. Of the 17 sites listed in Table 2,
ten were used only for ﬁsh collections, two were used only for drift collections, and ﬁve
were used for the collection of both ﬁsh and invertebrate drift. The Thompson’s Orchard
site (RM 239) was used for drift collection in 1995 only, whereas the western and eastern
shores of the Above Red Bluff Diversion Dam (ARBDD) site (RM 243) were used for
drift samples in 1996 only. The Red Bluff Research Pumping Plant (RBRPP) (RM 242)
and California Department of Fish and Game (CDFG) rotary screw traps (RM 278) were

used as ﬁsh collection sites in 1996 only.

 

i 7 1.2
{:1 rm '8‘: 2” E

RH 283 i
f E.»- mania/em...

(ottoman F!” Jr

Shasta County M N 3
rev .5 1.3:.

 

        
    
 

 

 

 

 

 

 

MI 258

Iolmrra County _ F
‘; RH 24‘ 4‘5
BED FIJI” 3 mg :43 g

Red Bluff a" 242

. RH 23!

amen-us m :50 an 33‘
. ms mums a:
1- | a
B 1-
" COIIMG n“ 2',

Butte County I'Ill LI F: 2
Ii. 5'20; .‘oo
Glenn County I ' I!" 193 “I 0 :1, I 3}

. , onuu ..
122.8”11 122.6"11 122. 4 N 122. 2”" 122.0”14 121.8 N 121.6 N

0 5 10 15 28 25 3801‘
Scale 1:85:88?" 9 m 29 39 4a 591'...

 

 

 

 

 

 

LEGEND
m Sacramento River 0 City marker
————lnterrtate 5 RH 26? Study Silo l'larlror
HM County Irorrlor (River Niles)

 

Figure 1. Map of the study area and location of study sites on the Sacramento River, 1995-
1996. Inset: Map of California showing location of the city of Red Bluff. Redrawn from the
US. Bureau of the Census, Tiger Mapping Service Database, 1997.

10

Table 2. Sites used for drift and ﬁsh sampling on the upper Sacramento River, California,
1995-1996. Sites are listed in order from downstream (Bidwell Park) to upstream
(Caldwell Park).

 

 

 

SITE COUNTY “if? 33%. sighs

Bidwell Park“ Butte 193 «I
Woodson Bridge“ Tehama 219

Sacramento Bar“ Tehama 236 \l \/
Thompson’s Orchards Tehama 239 ‘1

Red Bluff Research Pumping Plant Tehama 242 i ‘1
Below Red Bluff Diversion Dam“ Tehama 242 ‘1
Above RBDD—West Side Tehama 243 v v
Above RBDD-East Side“ ‘ Tehama 243 «J 4
Pizza Place“ Tehama 246 \I ‘1
Bend Bridge“ Tehama 258 \l ‘1
Table Mountain Tehama 267 «I

Mouth of Battle Creek“ Tehama 272 ‘1
Balls Feny“ Tehama 276 ‘1
Rotary Screw Traps" Tehama 278 ‘1
Anderson River Park“ Shasta 283 ‘l
Posse Grounds“ Shasta 298 \l
Caldwell Park“ Shasta 299 ‘1

 

*Denotes US. Fish and Wildlife Service, Northern Central Valley Fishery Resource Ofﬁce (Red
Bluff, California) seining sites.

"Operated by the California Department of Fish and Game, Environmental Services Division, Red
Bluff, California.

11
Drift Collection and Handling

Moﬂecm of Invertebrate Drift. Food abundances were estimated using replicated
drift nets (up to four) arranged in a transect set perpendicular to the river’s edge.
Sampling was repeated throughout the day at a given site. Most drift samples were
collected between sunrise and sunset, to compare with salmon feeding habits data which
were primarily collected during the day. Safety concerns and access restrictions at night
also precluded extensive nighttime sampling. Drift was sampled at one site per week, not
concurrently with ﬁsh sampling efforts except on one occasion (a 24-hour series; see
below).

Two types of nets were used, that differed in mouth dimensions and ability to
sample at different depths. The ﬁrst had a mesh size of 260 um, a removable bucket at the
closed end, and mouth dimensions of 45 x 30 cm (Wildlife Supply Company, Saginaw
MI). These nets could be oriented with either the long or short side parallel to the
substrate, and were most appropriate for shallow near-shore sampling. The second net
type had a 75 x 15.5 cm mouth and 64 um mesh, and were most often used in deeper
water further ﬁ'om shore, where the shorter nets could not capture surface drift.

With the drift net mouths facing into the current, nets were anchored into the
substrate with steel rods. In 1995, drift nets sampled the water for 30 to 60 min. In 1996,
I reduced sampling duration to 15 to 25 min, which improved net efﬁciency by reducing
net clogging and backwash. The start and stop time, water depth, and current velocity
(m/s) for each net were recorded for later use in calculations of drift density (number of
organisms per unit volume of water ﬁltered by the net). Water turbidity, dissolved

oxygen, and temperature were also recorded.

12

After removal of a net from the water, the contents were washed down to the
closed end of the net. The sample was sieved to remove excess water and transferred to a
plastic bag or bottle, where a solution of 10% formalin was added as a preservative. One
of two biological dyes, rose bengal or eosin-b, was also added to each sample to facilitate
sorting of organisms.

Processing of Drift Sar_nplea. In the laboratory, each drift sample was sieved and washed
to remove excess formalin. To expedite processing, some samples were suspended in
water and divided into subsamples using a Folsom plankton splitter (Wildlife Supply
Company, Saginaw, Michigan). The plankton splitter uses a rotating drum equipped with
a dividing blade to separate a liquid sample into two subdivisions of equal volume, which
are then poured into receptacles, or “boats.” The contents of one boat were returned to
the drum to be divided again if smaller subsamples were required. In order to identify a
minimum of 100 organisms per sample, sorting ranged ﬁ'om an entire sample to as little as
1/64 of a sample.

The sample or subsample was viewed under a lighted magnifying lens.
Invertebrates and ﬁsh larvae were sorted from the sample and transferred to a bottle of
80% ethanol solution. Sorted organisms were viewed with a microscope, counted, and
identiﬁed using published keys (Merritt and Cummins 1984, Pennak 1978, Usinger 1956,
Amett 1985). Drift density was estimated with the equation: DD = 100 x ((Number of
organisms per net hour)/(m3 ﬁltered per net hour)) (Allan and Russek 1985). The number

of organisms per net hour was calculated after adjusting for subsample size.

1 3
Fish Collection Methods

Beach Semg' ' . Beach seines (1.21 x 9.09 to 22.73 m; 3.2 mm mesh) were the primary
gear used to collect juvenile chinook salmon. With the exception of one diel series (see
below), all beach seining was conducted in coordination with the US. Fish and Wildlife
Service, Northern Central Valley Fishery Resource Ofﬁce, Red Bluff, California
(USFWS). Each USFWS site was sampled between 0800 and 1600 hours once every two
weeks barring safety or access problems. Sampling began at the Bidwell Park site (RM
193) and proceeded upstream over a two-day period. Sites were gravel bars, boat ramps,
or included both types of habitat. One or two seine hauls were made at each USFWS
gravel bar site. A second haul was made (upstream of the ﬁrst) if less than 50 juvenile
salmon were captured during the ﬁrst haul. A single seine haul was made at each boat
ramp.

Fish were removed ﬁ‘om the seine and held in a tub of ﬂesh river water for
processing. Juvenile salmon were anaesthetized prior to measurements using tricaine
methanesulfonate (MS-222) mixed with river water. The fork length of each ﬁsh was
measured to the nearest millimeter, and the run membership (fall, late-fall, winter, or
spring) of each chinook juvenile was estimated based on ﬁsh length and date of capture
using the run identiﬁcation model developed by Fisher (1992) and Greene (1992). After
length measurement, each juvenile salmon was blotted with a soft cloth and weighed to the
nearest 0.1 g (in 1995) or 0.01 g (in 1996).

During April-June 1996, up to 30 fall-run salmon per site were sacriﬁced by
prolonged exposure to the anaesthetic, for use in diet and percent water analyses.

Sacriﬁced ﬁsh were preserved by one of two methods: some ﬁsh were placed in bags of

14
water kept in an ice-ﬁlled cooler and later frozen (salmon used for estimating body water),

the rest were preserved in 10% formalin (salmon used for diet analysis). Many ﬁsh kept in
the cooler were measured several hours after sacriﬁce due to time constraints. Appendix
E details the experiments used to develop correction factors for length and weight
measurements of ﬁsh held in the cooler.

Other Meth_ods of Fish Capture. Two additional methods of capture other than beach
seining were used to collect information on juvenile chinook salmon. Live lengths and
weights of salmon entrained at the US. Bureau of Reclamation Red Bluff Research
Pumping Plant (RBRPP) (RM 242) were recorded during April 1996. Live measurements
and samples of fall-run juveniles were also collected at the CDFG rotary screw traps (RM
278) once per month during April-June 1996. Most of the salmon collected by
entrainment or in screw traps were captured at night and processed in the morning. The
methods of handling juvenile chinook after capture were the same as the methods

described for beach seining.

Salmon Diet Analysis
Juvenile salmon captured during April-June 1996 were examined to determine the
number, taxonomic composition, and total wet weight of stomach contents. Fish were
measured to the nearest millimeter (fork length), blotted, and weighed to the nearest 0.001
g. Dial calipers were used to measure the gape width of each specimen to the nearest 0.05
mm. The stomach was then excised and blotted, and the stomach weight, with and
without the contents, was determined to the nearest 0.00001 g with a Sartorius Research

R160D semi-microbalance. The total wet weight of the contents was calculated by

15
subtraction. A stomach fullness index (SFI) (HySIOp 1980) for each ﬁsh was calculated

with the equation:

SFI (%) = 100 x {(Total wet weight of stomach contents)/(Total ﬁsh wet weight)}
Stomach contents were viewed under a microscope ﬁtted with an ocular '
micrometer. Each organism was identiﬁed to the lowest practical taxonomic level (family

in most cases), and its life stage (larva or nymph, pupa, or adult) was recorded when
possible. The ﬁrst ten representatives encountered from each invertebrate taxon were
measured for head capsule width (HCW) to the nearest 0.03 mm, for organisms with a
head capsule. Head capsule widths were used to represent relative invertebrate prey size
since fragmentation of prey during swallowing and digestion precluded the use of other
measures, such as prey length, in stomach analyses. Total lengths of ﬁsh larvae found in
the stomach were measured to the nearest millimeter if the condition of the specimen
permitted. Copepods, cladocerans, nematodes, and aquatic worms were counted if
present, but were not measured. A large number of larval aquatic mites (Hydracarina)
were present in the contents of some ﬁsh; these were noted but not counted as prey items
because they were parasites on chironomid (Diptera) pupae, the intended prey of the
salmon. Adult mites were counted as prey. Nematodes, which could not be distinguished
as ﬁsh parasites versus recently consumed items, were enumerated as prey. Fragments of
organisms without a head attached (legs, wings, or abdomens) were neither counted nor
identiﬁed; it was assumed that the source organism had already been represented in the
count of heads or head capsules.

The linear selection index (L) (Strauss 1979) was used to compare diet with drift.

This index is described by the equation L = r.- — p,- , where r,- is the relative abundance

16
(expressed as a proportion) of taxon i in the diet of chinook salmon, and p.- is the

proportion of taxon i in the drift during the same time period. The index ranges from +1
to —1 , with positive values indicating preference, negative values indicating avoidance or
inaccessibility, and values of zero indicating random feeding (Strauss 1979). Selection of
individual taxa by juvenile chinook was calculated from abundance in the drift, totaled
across all sites and sampling dates from April-June 1996, compared with the abundance in
the diet of all ﬁsh examined from the same time period.
Drift ar_1_d Feedipg over the Diel Cycle. A series of coordinated drift and ﬁsh samples were
collected to compare diel patterns of feeding intensity and size selection of juvenile
chinook salmon in the Sacramento River with the density and size distribution of the drift
during the same 24-hour period. The study took place at the ARBDD-West site, a gravel
bar located between the mouth of Red Bank Creek (upstream boundary) and the Red Bluff
Diversion Dam (just downstream, at RM 243). A 1.21 x 9.09 m beach seine (3.2 mm
mesh) was used to collect juvenile chinook salmon at four-hour intervals. Sampling
commenced at 2000 hours Paciﬁc Daylight Time (PDT) on 19 April 1996 and ended at
2000 hours on 20 April 1996. Hauls were made from upstream to downstream. In some
cases, several hauls were required to capture ten juvenile chinook salmon conforming to
fall-run size criteria at each time step. Successive seine hauls progressed upstream of
previous hauls during a given sampling interval. Collected salmon were measured and
sacriﬁced using the same procedures described in previous sections.

A transect of three drift nets was used to sample the river four times during the 24-
hour period, at times intervening between the collection of ﬁsh samples. Samples from

two nighttime transects and two daytime transects were used to estimate nocturnal and

17
diurnal food size and taxonomic composition. The size (head capsule width, HCW)

distribution of the drift was obtained by pooling organisms from the three nets in a
transect. The pool was subsampled to produce about 100 organisms, which were
measured to the nearest 0.03 mm with an ocular micrometer.

Total 24-hour consumption by juvenile chinook salmon was calculated by
estimating consumption within each sampling interval, following the method of Sagar and
Glova (1988):

c. = {(s. — Soe'R‘)Rt}/(l — 6"“) c2. = z c.
Where Q = amount of food (wet weight, g) consumed in time t
S. = amount of food in stomach after t hours

S0 = initial amount of food in stomach

R = exponential rate of gastric evacuation (0.152 at 14°C, estimated by Kolok
and Rondorf 1987 for chinook salmon in the ﬁeld)

e = base of natural logarithms

C24 = amount of food consumed over the 24-hour period.

Condition Estimates
Condition ﬁctor. The lengths and weights of juvenile chinook salmon collected in the
ﬁeld were used to calculate the condition factor (K) for each ﬁsh, with the equation:

K = 105 x {Wet weight (g)/ Fork length (111303}

Dry Weight and PercenL Water. Juvenile salmon collected in the ﬁeld and preserved
frozen were sent to the Michigan State University Fisheries Laboratory, East Lansing,
Michigan, for drying. Fish were thawed, and their fork lengths (nearest mm) and weights
(nearest 0.001 g) were recorded. Fish were cut into halves or fourths and placed into

aluminum drying pans bearing unique identiﬁcation numbers. The ﬁsh were then dried to

18
a constant weight in a drying oven at 70°C. The dry weight of each ﬁsh was recorded to

the nearest 0.001 g. Percent water was estimated with the equation:
Percent Water = 100 x {(TW — DW)/T W}

where TW = the wet weight (g) of ﬁsh after thaw, and DW = dry weight (g).

Statistical Methods

Least squares regression lines of the form y = a + bx were ﬁtted to ﬁsh length-
wet weight and length-dry weight data, where x is ﬁsh length, y is ﬁsh weight, a is the
intercept and b is the regression slope. The data were transformed using natural
logarithms to linearize the relationships and to stabilize error variance. Least squares
methods were also used for regressions of ﬁsh gape width and prey head capsule width
(HCW) (mean, maximum, and minimum) on fork length. For fork length-HCW
regressions, only ﬁsh with more than ten items in the stomach were used, to reduce
variability due to low feeding intensity and low selection. A chi-square test of
homogeneity was used compare the size (HCW) distributions of invertebrates in the drift
during daytime and nighttime for the 24-hour series, and to compare the size distribution
of the 24-hour drift with the 24-hour diet. Length, weight, condition factor (K) and
percent water of ﬁsh captured in the ﬁeld were analde in relation to month and river
mile of capture using the SAS general linear models function (GLM), which performs a
generalized analysis of variance appropriate for unbalanced data (SAS Institute, 1989).
Length-wet weight and length-dry weight relationships of samples pooled by month and
river mile of capture were compared with the slope heterogeneity test and armlysis of

covariance (Littell et a1. 1991). Fish smaller than 50 mm FL and greater than 90 rmn FL

19
were excluded from length-weight slope heterogeneity tests since they affected the

linearity of the regressions. Unless otherwise stated in the results, a signiﬁcance level of a

= 0.05 was used in statistical tests.

RESULTS

Invertebrate Drift and Salmon Diet
Drift Density. Throughout the study, mean drift density ranged from 211 to 7214
organisms per 100 m3 (Table 3), with an overall average of about 771 organisms per 100
m3. The highest estimate originated from a set of samples taken after dusk during a 24-
hour series at the ARBDD site on 19 April 1996, with a mean of 7214 organisms/ 100 m3 .
The lowest estimates were from Bend Bridge during April and June 1996 (21 l
organisms/ 100 m3). Sacramento Bar (RM 236) and Bend Bridge (RM 258) appeared to
have the most stable drift densities across time, whereas ARBDD (RM 243) and Table
Mountain (RM 267) exhibited the most variability in mean drift density. Table 3 reveals
no consistent trend in drift densities from upstream to downstream sites.
Ralative Abundances of Drift Taxa. The drift was dominated by aquatic taxa, which
comprised between 96 and 99% of organisms in the drift during the three seasons studied
(Table 4). The most prevalent members of the drift were larvae, pupae, and adults of the
family Chironomidae (Diptera), which comprised 69 and 50% of the total number of
organisms collected during July-August 1995 and April-June 1996, respectively, and 20%
during January-March 1996 (Table 4; see Appendices A and B for detailed taxonomic

information). The high abundance of chironomids was responsible for the high general

20

Table 3. Mean, standard deviation, and 95% conﬁdence for drift density (number of
organisms per 100 m3 water ﬁltered per net) by site and date for the upper Sacramento
River, 1995-1996. RM = river mile. N = number of samples (nets) taken each date.

 

Approximate

 

Mean Drift Standard
SITE RM DATE Deviation C fd
(No. per 100 m3) 0'” 6"“
Sacramento Bar 236 7/22/95 6 673 361 289
7/29/95 12 703 758 429
2/25/96 10 549 691 429
5/5/96 1 5 363 249 126
6/9/96 12 538 433 245
Thompson’s 239 7/13/95 3 499 263 298
8/9/95 12 838 501 283
Above RBDD 243 1/23/96 3 3108 1047 1 184
1/24/96 6 1498 969 776
3/3/96 6 390 213 171
3/21/96 12 676 210 119
4/19/96“ 3 7214 9714 10992
4/20/96“ 9 465 217 142
5/13/96 14 670 752 394
6/23/96 9 2100 3576 2337
Pizza Place 246 7/25/95 12 712 326 184
1/30/96 1 l 281 75 45
5/3/96 2 483 227 315
6/23/96 2 1537 41 1 570
Bend Bridge 258 7/1 U95 8 425 179 124
8/6/95 16 594 455 223
2/6/96 8 231 131 91
3/19/96 12 243 I68 95
4/14/96 11 211 114 68
6/2/96 10 21 l 102 63
Table Mountain 267 7/26/95 12 1498 623 352
8/16/95 3 4012 3795 4295
2/11/96 5 361 216 189
4/7/96 6 805 661 529
5/19/96 11 638 559 331
6/30/96 9 295 163 106

 

* Estimates include samples taken at night. All other estimates include only samples taken between

sunrise and sunset.

22
contribution of the order Diptera. The family Baetidae (Ephemeroptera), particularly

nymphs, also contributed large numbers to the drift, with 9 to 17% of the totals. The
“other” category, made up of such items as copepods and oligochaetes, was responsible
for between 5 and 47% of the drift in all sampling periods. The order Plecoptera -
represented less than one percent of drift numbers except during January-March 1996,
when it made up almost 10%. When percentages were recalculated excluding the “other”
category (Table 4) to reduce the effects of differential sorting of samples by different
workers, the percentage of dipterans was still low in January-March 1996 (47%), due to
the greater relative abundance of plecopterans during that season.

Relative Abundances of Taaa in the Die_t. The diets of 189 juvenile chinook salmon were
examined. As with the drift, the diet was dominated by aquatic taxa, which made up 94%
of all items consumed. Chironomids of all life stages were the predominant prey, making
up over 60% of the diet during April-June 1996 (Table 4; see Appendix B for detailed
diet information). Diptera as an order made up about 67% of the diet, mostly due to the
high contribution of chironomids. Baetid nymphs were the next most common prey item,

with 14% of the total. Trichopterans, homopterans and items from the “other” category

were also found in modest numbers in salmon stomachs.

23

Table 4. Percentages of common taxa found in the drift and diet of juvenile chinook
salmon in the upper Sacramento River during 1995 and 1996 (n = number of drift or
stomach samples examined, n. = total number of items identiﬁed and used to estimate
abundances). Drift percentages recalculated without the “other” category are included in
parentheses (see text for explanation).

 

 

TAXON/LIFE STAGE JULY-AUG. JAN-MARCH APRIL-JUNE APRIL-JUNE
1995 DRIFT 1996 DRIFT 1996 DRIFT 1996 DIET
(n=84, (n=74, (n=113, (n=189,
n. = 12,264) n. =7768) n. = 11,567) n. =7112)
Diptera Total 75-55 (79-89) 25-03 (47-13) 55.27 (70.03) 67.08
Chironomidae Larva 51-43 1632 37.12 20.25
Pupa 6'03 0'93 2.08 12.98
Adult 1 1-68 2-70 1 1.28 28.95
Coleoptera 0.21 (0.22) 0.66 (1.24) 0.65 (0.82) 0.38
Ephemeroptera Total 16.40 (17.34) 11.21 (21.12) 17.63 (22.34) 17.51
Baetidae Nymph 14.91 8.88 15.79 14.15
Adult 1.10 0.21 1.15 2.40
Hemiptera 0.77 (0.82) 0.15 (0.29) 0.43 (0.55) 0.21
Homoptera Total 0.61 (0.65) 1.22 (2.30) 1.20 (1.52) 3.12
Aphididae Nymph 0.1 1 0.08 0.03 -
Adult 0.34 0.10 0.61 2.35
Hymenoptera 0.21 (0.22) 0.30 (0.56) 0.92 (1.16) 1.38
Plecoptera 0.08 (0.09) 9.78 (18.42) 0.18 (0.23) 0.39
Trichoptera Total 0.57 (0.60) 3.10 (5.84) 1.29 (1.63) 4.01
I-lydropsychidae Larva 0.38 1.15 0.41 1.98
Adult - - 0.04 0.84
Collembola 0.05 (0.05) 1.08 (2.04) 0.97 (1.23) 0.03
Other insects 0.11 (0.12) 0.57 (1.07) 0.39 (0.49) 0.23
Other 5.42 46.90“ 21 .08 5.65
Terrestrial Total 1.40 2.57 3.54 5.62
Aquatic Total 98.60 97.43 96.46 94.38

 

*Primarily oligochaetes, calanoid and cyclopoid copepods, and aquatic mites (Hydracarina). The
high percentage of the “other” category for January-March 1996 may be due to more thorough
sorting of these samples.

24

F requen_cy of Occurrence in the Diet. Frequency of occurrence, or the percentage of ﬁsh
which consumed one or more representatives of a taxon, conﬁrms a general pattern of
feeding upon taxa abundant in the drift. Table 5 shows that organisms present in most of
the 189 stomachs included chironomids of all life stages (larvae—75%, pupae—60%,
adults—80%), baetid nymphs (53%), and hydropsychid larvae (33%). Most other taxa
and life stages were present in less than 10% of the stomachs; these were excluded ﬁom
Table 5.

Selection of Available Prey by Juvenile Chinook. Selection indices were calculated to
quantify the preferences for, or accessibility of, common prey taxa during April-June 1996.
For most groups, the linear selection index (L) was close to zero, suggesting random
selection of those items. Values of L ranged only ﬁom —0. 1 7 to a maximum of 0.18
(Figure 2). Selection was positive for the orders Diptera (0.12), Trichoptera (0.03), and
Homoptera (0.02), and negative for Collembola (—-0.01). Among the taxa numerically
common in the drift, negative values of L were observed for chironomid larvae (—0. 17),
baetid nymphs (—0.02), and simuliid larvae (—0.02), whereas positive values were seen for
baetid adults (0.01), chironomid pupae (0.11), and chironomid adults (0.18).
Oligochaetes, an important component of the drift, were negatively selected in the diet
(—0.05). Other small items numerically abundant in the drift were also negatively selected,

including copepods (-0.05) and cladocerans (—0.04).

25

Table 5. Frequency of occurrence (number of stomachs containing one or more items of a
taxon) of selected prey taxa of juvenile chinook salmon captured from the upper mainstem
Sacramento River, April-June 1996. Only items present in over 10% of stomachs are

listed.

 

 

Number of Stomachs F ‘ f
Taxon Life Stage with Taxon Present requency 0
_ Occurrence (%)
(n - 189)
Diptera
Chironomidae Larva l 42 7 5
Pupa 113 60
Adult 151 80
Simuliidae Larva 55 29
Adult 28 15
Empididae Adult 21 1 l
Ephemeroptera
Baetidae Nymph 1 01 53
Adult 36 19
Homoptera
Aphididae Adult 52 28
Trichoptera
Hydropsychidae Larva 63 33
Adult 24 13
Unidentiﬁed Adult 1 9 1 0
Hymenoptera
Unidentiﬁed Adult 30 1 6
Teleostei Larva 37 20
Nematoda Unidentiﬁed 3 3 l 7
Arachnida Unidentiﬁed 22 1 2

 

26

 

 

 

 

 

 

$803.0
88800
23:89.0
28.5.80
«23 325836»:
805055.
$238.6.
82822:»:
.32 822:?
82350:

Taxon

meager

=33. oauzomm
$53 $238
Soaoeoeocaw
8058.00
«23 32.356
:32 3282.820
maze omEEocero
«23 3288820
8220

 

 

 

 

ll
4 .— d «I
5
n2. 1. m m. o m m .4... m
o o o o

3. x00... c2832..“ .35..

Figure 2. Linear selection indices (L) for selected taxa found in the drift and

the diet of juvenile chinook salmon in the upper Sacramento River, April-

June 1996. Possible values of L range ﬁom 1 to -l.

27

IE1 Size. Gape Size and Prev Size Ralatioﬂlips. The gape width of juvenile chinook
salmon showed a signiﬁcantly positive linear relationship with fork length (y = 0.0859x —
0.5405; r2 = 0.82, P < 0.0001, n = 189) (Figure 3). The relationship between fork length
of juvenile chinook and maximum head capsule width (HCW) of measurable invertebrate
prey was signiﬁcantly positive (t2 = 0.33, P < 0.0001, n = 134) (Figure 3). The
relationship between fork length and mean HCW was also positive (r2 = 0.26, P < 0.0001,
n = 134). The relationship between fork length and minimum HCW was signiﬁcantly
diﬂerent from zero, but the correlation was weak (r2 = 0.05, P < 0.01, n = 134). The
mean gape width of all ﬁsh examined was 5 mm (iSE = 0.09; ranging from 2.05 to 8.15
mm; 11 = 189), whereas the mean HCW of all measurable invertebrate prey items was 0.55
mm (iSE = 0.01; ranging from 0.06 to 3.60 mm; 11 = 3742). Table 6 describes the size
ranges of representative prey taxa from the diet.

Fish larvae did not comprise a large percentage of the total number of items in the
diet; and the size of ﬁsh prey, measured in length, cannot be compared directly with the
sizes (HCW) of invertebrate prey. However, there was a trend in the relationship between
the size of juvenile chinook and the number of ﬁsh larvae consumed. Of the salmon that
consumed ﬁsh larvae (n = 37), all were 40 mm FL or larger. Only juvenile chinook of at
least 60 mm FL had stomachs containing ﬁve or more ﬁsh larvae (the maximum was 38).

Fish larvae in salmon stomachs ranged from 3 to 19 mm total length.

Gape Width or Prey Head Capsule Width (mm)

28

9.00

 

  
  
  
 
 

800 4 Y = 0.0859X - 0.5405 lo Gape WIEIII I o o
R2 = 0.8162 l. Max How 1 o
y = 0.0243x — 0.331 l0 Mean HCW 3 0 0°
2 = ,. Min ch
7.00 « R 0.3266 1 - 7 7 9
y=0.0061x+0.1587 ° 6 o
R2 = 0.2604 6 o 0,, 8
o o °°°
6.00 « y = 0.0014x + 0.1442 0 0° 03 o e

R2 = 0.0499

 

 

 

 

 

Fork Length (mm)

Figure 3. Regressions of gape width, and maximum, mean, and minimum head
capsule width of invertebrate prey, on fork length of juvenile chinook salmon
captured April-June 1996 from the upper Sacramento River.

29

Table 6. Mean, standard deviation, maximum, and minimum head capsule widths of
common taxa found in stomachs of juvenile chinook salmon captured in the upper
Sacramento River, April-June 1996.

 

 

Taxon Life Stage N0. measured Mean HCW SD (i) Min, Max
Chironomidae Larva 668 0.28 0.08 0.09, 0.63
Chironomidae Pupa 590 0.47 0.08 0.21, 0.87
Chironomidae Adult 974 0.45 0.08 0.15, 0.84
Baetidae Larva 411 0.75 0.17 0.24, 1.11
Baetidae Adult 71 0.84 0.11 0.60, 1.17
Simuliidae Larva 143 0.47 0.12 0.18, 0.81
Aphididae Adult 128 0.42 0.06 0.27, 0.60
Hydropsychidae Larva 133 1.13 0.36 0.39, 2.22
Hydropsychidae Adult 54 1.83 0.18 1.11, 2.25

 

Feeding Intensity and Stomach Fullness. The number of prey items per stomach and
stomach fullness index (SF 1) were used as measures of the feeding intensity or energy
intake of juvenile chinook salmon in the Sacramento River. Only two of the 189 ﬁsh
examined had empty stomachs. The SFI averaged about 2.3% (iSD = 1.6) and ranged
from 0 to over 8% of wet weight. Arrangement of data into three size classes and three
monthly categories rendered no clear trends in feeding intensity, as measured by mean SFI
or mean number of prey per stomach (Table 7). During April, an increase in mean number
of prey per stomach with increasing size was apparent, while the highest mean stomach
fullness was seen in the smallest size class (30 to 50 mm). However, in June, both the
mean number of prey and the mean SFI increased with ﬁsh size.

Driftﬁand Feeding over the Diel Cycle. The sun set at approximately 1952 hours PDT on
19 April 1996, rose at 0623 hours on 20 April, and set at 1953 hours on 20 April (US.
Naval Observatory, unpublished data). Average daily water temperature at Red Bluff was

10.7°C on 19 April and 102°C on 20 April (US. Bureau of Reclamation, unpublished

30

Table 7. Mean number of prey per stomach and mean stomach fullness index (SFI) of
juvenile chinook salmon captured in the upper Sacramento River, April-June 1996,
arranged by month and size class, and presented with approximate 95% conﬁdence limits.

 

 

Approx. Approx.
Month Size Class Sample Mean No. 95% Mean SH 95% .
(mm) Size (n) of Prey Conﬁdence (%) Conﬁdence
(i) (3%)
April 30-50 29 24 10 3 .0 0.8
51-70 40 34 10 2.1 0.3
71-91 35 4O 15 2.5 0.7
May 30-50 3 29 27 1.8 0.5
51-70 34 33 11 2.4 0.5
71-91 18 26 10 1.6 0.4
June 30-50 3 4 4 1.1 1.1
51-70 4 31 24 1.7 0.2
71-91 23 80 26 2.2 0.3

 

data). Light precipitation occurred during the evening of 19 April. The gates of the Red

Bluff Diversion Darn were raised, so the site was characterized by riverine rather than

reservoir conditions during the study. Turbidity readings taken at the time of each seine

haul indicated relatively clear water, with measurements ranging from 3.48 to 6.75 NTU

over the 24-hour period.

The number of ﬁsh captured was highest near sunset on both days of sampling. At

0415 hours, two juvenile salmon were captured; only one of these ﬁsh was designated as

fall-run and could be sacriﬁced. From three to ﬁve ﬁsh were examined from all other time

periods. The size of juveniles examined for the study ranged from 38 to 79 mm FL (based

on measurements from formalin-preserved specimens).

31
The relationship between feeding intensity and estimated drift density over the 24-

hour period is shown in Figures 4 and 5. Mean drift density peaked (7214 organisms/ 100
m3) one hour after sunset on 19 April and remained near 500 organisms/ 100 m3 for the
other sampling periods. Mean stomach ﬁillness index (SF I) peaked two hours before
sunrise on 20 April, with a higher peak approximately three hours before sunset on 20
April (Figure 4). Mean number of prey per stomach peaked at sunset on 19 April,
declined throughout the night on 20 April, and peaked again near sunset on 20 April
(Figure 5). Total consumption over the 24-hour period was estimated as 12.5% of ﬁsh
wet weight.

The size distribution of the invertebrate drift was not constant over the 24-hour
period, with daytime drift signiﬁcantly different from night drift (x2 = 65.7, df= 9, P <
0.0001). Night drift included more large items. The overall size distribution of the drift
during the 24-hour period was also signiﬁcantly different ﬁ'om the size distribution of
invertebrates in salmon stomachs (x,2 = 134.5, df= 12, P < 0.0001) (Figure 6). The
average size of prey was close to 0.5 mm in both the drift and diet, but juvenile salmon

included larger prey sizes in their diet than were commonly available in the driﬁ.

Mean Drift Density (#I100 m3)

8000

32

 

7000 -.

6000 --—

5000

4000

3000 .

2000

1000 4

 

Sun set just before hour 0
Sunrise at hour 10.38

+_Drift Density i
_ 1

—0— SF1

Sunset at hour 23.

 

 

0 4 8 12 16 20

Hour of Capture

24

<~ _4.5

3.5

2.5

~ 1.5

0.5

Figure 4. Mean driﬁ density (number of organisms per 100 m3) and stomach
fullness index (SFI) (% of total ﬁsh wet weight) of juvenile chinook salmon

captured from the Sacramento River, 19-20 April 1996. Sampling began at 2000
hours PDT (hour 0).

(7.) xepul sseuung qaemors ueew

33

 

 

 

8000 100
Sun set just before hour 0 Sunset at hour 23.8;
Sunrise at hour 10.38
‘ 1* 90

7000 .

«~ 80
6000 .

«r 70
5000

 

'+"“"onn6;n;ny‘ * ~~6o
—o—Mean No._of Prey

\ 1 /1:
m \\

 

 

Mean Drift Density (#1100 m3)
‘S E

uaewois Jed KOJd Jo JaqlunN ueaw

 

 

 

 

1* 20
1000
W ‘4’ 1o
0 l I I l I o
0 4 8 12 16 20 24

Hour of Capture

Figure 5. Mean drift density (number of organisms per 100 m3) and mean number
of prey per stomach of juvenile chinook salmon captured from the Sacramento
River over a 24-hour period, 19-20 April 1996. Sampling began at 2000 hours
PDT (hour 0).

Frequency

34

 

160

 

140

 

 

120 l IDiet

 

1001

 

 

60

 

 

 

40 —

 

20«

 

 

 

 

 

 

 

 

‘b o '\ g '5 N e

N 0?? 9?? 61 $6 ,9? 49' :5 35° .3 ,9 $4,;
HCW Class (mm)

Figure 6. Frequency distribution of invertebrates of different size classes (head

capsule width) present in the drift and juvenile chinook salmon diet over a 24-
hour period, 19-20 April 1996, in the Sacramento River.

35

Taxonomic composition of the drift also differed between night and day during the
24-hour period (Table 8). Baetid nymphs made up over 60% of the nighttime drift but
only 22% of the daytime drift. Chironomid larvae and adults dominated the daytime drift,
making up about 42% combined. Terrestrial taxa, such as aphids, and semiaquatic taxa
such as collembolans and hymenopterans, were more abundant in the drift during the
daytime than the nighttime. Fish larvae (Teleostei) were captured in drift nets in greater

abundance during the night.

Table 8. Daytime and nighttime taxonomic composition of the Sacramento River drift
during one 24-hour period, 19-20 April 1996.

 

 

. Day(n=701) Night (n= 1168)

Taxon L'fe Stage Number Percent Number Percent

Chironomidae Larva 127 18.12 95 8.13
Chironomidae Pupa 24 3.42 1 3 l .1 1
Chironomidae Adult 165 23.54 1 16 9.93
Other Diptera All 25 3.57 9 0.77
Baetidae Nymph 152 21 .68 723 61.90
Other Ephemeroptera All 7 1.00 8 0.68
Plecoptera All 0 0.00 3 0.26
Aphididae Adult 20 2.85 0 0.00
Other Homoptera All 1 0.14 1 0.09
Trichoptera All 2 0.29 1 0.09
Coleoptera All 1 6 2.28 1 0.09
Hymenoptera Adult 8 1.14 0 0.00
Collembola All 32 4.56 0 0.00
Teleostei Larva 29 4.14 132 1 1.30

Other All 93 13.27 66 5.65

 

36

Condition of Juvenile Chinook Salmon
Le h-Wei t and Len h-Condition Relationshi s. The overall regression of In wet
weight on In length for juvenile salmon captured from the Sacramento River was described
by the equation y = 3.4874x - 6.6174 (r2 = 0.99, P < 0.0001, n = 1204) (Figure 7). The
apparent allometric length-weight relationship, characterized by a regression slope b
greater than 3.0, resulted in a curvilinear relationship between length and condition factor
(K) (Figure 8), with values of K increasing with an increase in length until a decline at
sizes greater than 80 mm. Figure 9 compares the overall length-weight regression line for
ﬁeld caught salmon (illustrated also in Figure 7) with regressions for experimental ﬁsh of
known two-week feeding history (experiments described in Appendix D).
mgth-Dry Weight Relationsjip. The In length-1n thawed wet weight relationship in
juvenile salmon sacriﬁced April-June 1996 was compared with the ln length-1n dry weight
relationship for the same ﬁsh (Figure 10). The regressions showed that ﬁsh dry weight (b
= 3.93) had a steeper slope than wet weight (b = 3.53) in relation to length, which
translated into a decrease in percent water estimates with increasing length.
General Linear Models Analysis of Spatial and Temmral Patterns. General linear models
(GLM) analysis indicated that the month and river mile effects on length, weight,
condition factor K and percent body water were signiﬁcant (P < 0.0001 across all
variables for both terms) with coefﬁcient of determination (r2) values of 0.36, 0.30, 0.39,
and 0.24, respectively (Table 9). The month x river mile interaction effects, though
signiﬁcant, were excluded from the analyses, since the absence of some month-river mile
combinations prevented meaningful interpretation. When included, interaction terms in

the models offered little extra ability to account for variation in the dependent variables.

37
Least squares mean length, weight, and condition factor K determined by the GLM

analyses generally increased from upstream to downstream sites, and from winter to
summer months (Tables 10 and 11). Percent water decreased over time but exhibited no
clear pattern with respect to river mile. Descriptive statistics for length, weight, condition
factor, and percent water (unadjusted by the least squares method) are presented in
Appendix C.

A__r;_alysis of Covarignce of Regressions by Month and River Mile. Because condition
factor and percent water results ﬁom the general linear models were strongly dependent
on mean length of ﬁsh sampled, the question remained whether weight or percent water
for a given length was different between months or sites. Slope heterogeneity tests
indicated that slopes of the ln length-1n weight regressions by month and river mile were
not signiﬁcantly different for the 50-90 mm size range, therefore analysis of covariance
was used to compare intercepts. Intercepts were signiﬁcantly different among months (P
< 0.0001) and sites (P < 0.0001). For the analysis of covariance, higher intercepts indicate
more weight (“better” condition) or a greater proportion of water (“worse” condition) for
a given length. Slopes of the length-percent water relationship were homogeneous
between months, but heterogeneous between sites (P < 0.0005). Tables 12 and 13
summarize the length-weight and length-percent water regression parameters by month

and river mile.

In Weight (mg)

38

11

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 4.
9
y = 3.4874x - 6.6174
R2=0.9909
8
7 m
6
O
5 1.
4 t 1 i i : t + t i
3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5 5.2

In Fork Length (mm)

Figure 7. Ln fork length (mm) vs. ln weight (mg) of juvenile chinook salmon
captured from the Sacramento River, California, 1995-1996 (n = 1204). Lengths
and weights have been adjusted for delayed measurements where appropriate.

Condition Factor (K)

39

1.60

 

1.40

 

1.20

 

 

 

 

1.00

0.80 .

 

 

 

 

0.60

 

0.40 9

 

 

0.20

 

«>—
:-

 

 

0.00 4
0 20 40 60 80 100 120 140

Fork Length (mm)

1 l J
l l T

_1._

Figure 8. Relationship between condition factor (K) and fork length of juvenile
chinook salmon captured from the Sacramento River, 1995-1996 (n = 1204).
Length and condition were adjusted for delayed measurement where appropriate.

160

ln Weight (mg)

40

 

 

 
 

 

 

 

 

 

11
10 f/i‘
..ri"'/ V
xii/M
9 y = 28558X - 38373 / y = 3.:874X - 6.6174
l R2 = (19202 .x’ R = 0.9909
y = 2.7797x - 3.6485 / x
8 ‘“ R2 = 0.914 , I .'
y = 2,22 _ 02;;5156 . ” F, . y = 4.0178x - 9.09
7 -. ' ’19?" R2 = 0.9389
6 1. y = 3.5225,, _ 7.0403 Linear (Exp.l Level 0)
R2 = 0,7416 — - - — Linear (Exp.l Level 1)
Linear (Exp. I Level 2) l
------ Linear (Exp.ll Level 0)
5 4 — — - Linear (Exp.ll Level 1)
WWW-- Linear (Field ) _
4 i i t i l i l
3.4 3.6 3.8 4 4.2 4.4 4.6 4.8

In Fork Length (mm)

Figure 9. Ln length (mm) vs. 1n weight (mg) of juvenile chinook salmon captured
from the Sacramento River, California, 1995-1996, compared with salmon from
grth experiments (Level 0 = unfed two weeks; levels 1 and 2, fed).

 

In Wet or Dry Weight (mg)

41

y = 3.5336x - 6.7763
R2 = 0.9513

DD
y= 3.9301x- 10.13
R2 =0.9422

1
l

in Dry Weight

0 Wet Weight

1

 

3.4 3.6 3.8 4 4.2 4.4 4.6
In Fork Length (mm)

Figure 10. Relationship between 1n fork length (mm) and In wet weight (mg) or In
dry weight (mg) for juvenile chinook salmon captured April-June 1996 ﬁ'om the

Sacramento River (n = 520).

42

Table 9. Results of general linear models analyses, describing the eﬂ'ect of month and
river mile of capture on length, wet weight, condition factor (K) and percent body water
of juvenile chinook salmon captured ﬁ'om the Sacramento River, 1995-1996 (n =1204 for
length, weight, and K; n = 520 for percent water).

 

 

Variable/Factor/ Sum of Mean . Coefﬁcient
Source df Squares Square F-ratio Pr > F r2 of Variation
Length
Model 16 140928 8808 41.09 0.0001 0.36 24.86
Error 1 187 254465 214
Corrected Total 1203 395393
Type 111 Main Effects
MONTH 5 88627 17725 82.68 0.0001
RIVERMILE 1 1 14842 1349 6.29 0.0001
Weight
Model 16 2526 158 31.06 0.0001 0.30 79.96
Error 1 187 6034 5.08
Corrected Total 1203 8559
Type 111 Main Effects
MONTH 5 1741 348 68.50 0.0001
RIVERMILE 1 1 247 22 4.41 0.0001
Condition Factor (K)
Model 16 13.36 0.83 47.93 0.0001 0.39 13.64
Error 1 187 20.68 0.02
Corrected Total 1203 34.04
Type 111 Main Effects
MONTH 5 8.13 1.63 93.38 0.0001
RIVERMILE 1 1 1.77 0.16 9.21 0.0001
Percent Water
Model 9 849 94.4 18.01 0.0001 0.24 2.79
Error 510 2672 5.24 '
Corrected Total 519 3521
Type 111 Main Effects
MONTH 2 519 259 49.51 0.0001
RIVERMILE 7 443 63.2 12.07 0.0001

 

43

Table 10. Least squares mean length, weight, condition factor (K), and percent water
(presented with standard errors) of Sacramento River juvenile chinook by month/year of
capture, averaged across all sites. The estirmtes were generated by analyses with general
linear models. Length, weight, and condition data based on delayed measurement were
adjusted to reﬂect live measurements (Live length = delayed length/(0.9844); live weight =
delayed weight/(10287)). Percent water was calculated using the thawed weights of ﬁsh
rather than the original live weights. N.a. = not applicable because ﬁsh were not sacriﬁced
during those months.

 

WElG HT (g) CONDITION

 

 

.. .2222:
July 1995 75 (2) 5.27 (0.31) 1.06 (0.02) n.a.
February 1996 44(1) 1.33 (0.18) 0.81 (0.01) n.a.
March 1996 47 (2) 0.99 (0.30) 0.92 (0.02) n.a.

April 1996 58 (1) 2.33 (0.16) 0.96 (0.01) 82.90 (0.16)
May 1996 65 (1) 3.29 (0.21) 1.06 (0.01) 81.06 (0.22)

June 1996 76 (2) 5.61 (0.26) 1.15 (0.02) 80.04 (0.36)

 

Table 11. Least squares mean length, weight, condition factor (K), and percent water
(presented with standard errors) of Sacramento River juvenile chinook by site (river mile),
averaged across months. Estimates were generated by analyses with general linear
models. Length, weight, and condition factor data based on delayed measurement were
adjusted to reﬂect live measurement (Live length = delayed length/(0.9844); live weight =
delayed weight/(10287)). Percent water was calculated based on the thawed wet weight
and dry weight of juveniles captured April-June 1996. Na. = not applicable, sample not
taken.

 

FORK CONDITION

 

RIVER MILE LENGTH (mm) WEIGHT (g) FACTOR (K) PERCENT
(:35) (iSE) , (i513) WATER (3:515)

193 66 (1 ) 3.88 (0.20) 1.05 (0.01) 81.52 (0.26)
219 66(1) 3.65 (0.20) 1.02 (0.01) 82.34 (0.29)
236 62(1) 3.13 (0.16) 1.02 (0.01) 81.53 (0.26)
242 65 (2) 3.37 (0.32) 1.03 (0.02) 79.45 (0.61)
243 68 (2) 4.41 (0.31) 1.05 (0.02) n.a.

246 62 (2) 3.08 (0.31) 0.98 (0.02) 80.46 (0.38)
258 58(2) 2.93 (0.32) 0.97 (0.02) 81.61 (0.63)
272 59 (3) 2.67 (0.43) 1.02 (0.03) 83.41 (0.46)
276 58 (6) 3.00 (0.96) 0.98 (0.06) n.a.
278* 59(1) 3.22 (0.17) 0.95 (0.01) 80.31 (0.20)
283 54(3) 2.23 (0.42) 0.95 (0.02) n.a.

298 54(3) 2.05 (0.42) 0.95 (0.02) n.a.

 

*Screw traps above Balls Ferry, the only site that involved gear other than beach seines.

45

Table 12. Slopes and intercepts for In length-1n weight regressions (for ﬁsh 50 to 90 mm
FL) and length-percent water regressions by month for juvenile chinook salmon captured
ﬁ'om the Sacramento River, 1995-1996. Pooled slope estirmtes were determined by slope
heterogeneity tests. Intercepts, determined by analysis of covariance, are ranked from
highest to lowest intercept for weight, and ﬁom lowest to highest intercept for percent
water.

 

 

 

 

Month Length-Weight Length-Weight Length-Water Length-Water
Slope or Pooled Intercept (a) Pooled Slope (b) Intercept (a)
Slope (b) (with rank for (with rank for

__ _ ANCOVA) ANCOVA)

July 1995 3.5362 —6.8393

February 1996 3.1621 —5.3448 (5)

March 1996 -5.2796 (4)

April 1996 —5.2385 (3) —0.1 123 89.3386 (3)

May 1996 —5.2131 (2) 88.4272 (1)

June 1996 —5.1583 (1) 88.5655 (2)

 

 

 

46

Table 13. Slopes and intercepts for In length-1n weight regressions (for ﬁsh 50 to 90 mm
FL) and length-percent water regressions by river mile for juvenile chinook salmon
captured from the Sacramento River in 1996. Pooled slope estimates were determined by
slope heterogeneity tests. Intercepts, detemiined by analysis of covariance, are ranked
from highest to lowest intercept for weight. Both length-water regression slopes and
intercepts are reported, as the slopes were found to be signiﬁcantly heterogeneous.

 

 

 

River Mile Length-Weight Length-Weight Length-Water Length-Water
Pooled Slope (b) Intercept (a) Slope (b) Intercept (a)
(with rank for
“___-_"_w ____ _'____________ ANCOVA) __ _ _

193 3.2263 —5.4834 (5) —0.1856 94.213

219 —5.4923 (6) —0.0880 88.716

236 —5.4730 (2) —0.1475 91.005

242 —5.5380 (9) -0.0886 88.032

243 -5.4798 (4)

246 —5.5473 (10) -0.1221 88.912

258 —5.5377 (8) —0.071 8 87.031

272 —5.4790 (3) -0.1017 89.984

276 -5.4481 (1)

278 —5.5003 (7) —0.1097 88.161

298 —5.604l (11)

 

 

DISCUSSION

Invertebrate Drift and Salmon Diet
Drift Density. Drift density provides a rough index of the productivity of a river. Average
drift density estimates in the Sacramento River, ranging from 211 to 7214 organisms per
100 m3, were comparable to estimates in other rivers, and higher than estimates obtained
for systems in which food is considered limiting for drift-feeding salmonids. Bowles and
Short (1988) found peak drift densities in a Texas stream to be near 1000 organisms/ 100
m3 in February, 600 in August, 500 in May, and below 100 in November. Drift density in
the Rakaia River, New Zealand, during freshwater rearing (a single 24-hour period) of
juvenile chinook salmon was between 200 and 900 organisms/ 100 m3 (Sagar and Glova
1988); the authors determined that, based on the size and condition of ﬁsh, food was not
limiting. In contrast, drift densities for several Southern Appalachian streams were as low
as 0.29 to 27.77 organisms/ 100 m3, resulting in food-limitation of resident salmonids
(Cada et a1. 1987).

The highest estimate of drift density in my study was recorded from night samples
at the ARBDD—West site, which is located just downstream of the mouth of Red Bank
Creek. Future studies could compare drift density of the main stem to that of the
Sacramento River tributaries, which have been shown to foster high grth rates in

juvenile salmon during February and March of wet years (Moore 1997). Such studies

47

 

48
might also consider what proportion of the main stem drift density is contributed by the

tributaries, and also whether a high concentration of food organisms at the conﬂuence of
tributaries could be the mechanism by which juveniles in the main stem are attracted into
tributary habitats.

RLlative Abund_ances of Tm Drift and Die_t. The taxonomic composition of the drift
and diet in the Sacramento River during 1995 and 1996 was similar to that described for
other river systems (Sagar and Glova 1987, Becker 1973) and to a previous study of the
upper Sacramento River (Schaftter et al. 1982).

As in other large rivers, drift and diet in the Sacramento River were dominated
numerically by a few common taxa with a small contribution ﬁ'om other groups. In the
Rakaia River, New Zealand, Deleatidium (Ephemeroptera: Leptophlebiidae) nymphs
comprised over 85% of the diet of juvenile chinook in the spring; Deleatidium made up
about 80% of the benthos, and 48% (daytime) to 70% (nighttime) of the drift, during
spring (Sagar and Glova 1987). During the summer, chironomids replaced Deleatidium as
the most common prey item, though the relative abundance of chironomids in the drift did
not change substantially from that of spring. In the central Columbia River, juvenile
chinook fed on chironomid adults (58 to 64% of the diet) and chironomid larvae (17 to
18%) (Becker 1973). Notonectidae (Hemiptera) and adult Hydropsychidae (Trichoptera)
were also locally important in the central Columbia, together making up around 5% of the
diet of juvenile chinook. Schaﬁter et al. (1982) determined that the diet of juvenile
chinook salmon in the upper Sacramento River during February-June 1981 was dominated

by chironomids, baetids, and aphids, similar to what was found in my study.

 

49

Organisms belonging to terrestrial taxa were relatively rare in the Sacramento
River drift in 1995 and 1996, comprising less than 4% in all seasons. Schaﬂter et al.
(1982) found the contribution of terrestrials (primarily Aphididae) to the Sacramento
River drift during spring to be as high as 50%. Aphids also made up as much as 38% of
the salmon diet in that study. The difference in the importance of terrestrials between the
two studies may reﬂect changes in riparian habitats over the past 15 years, such as
possibly increased use of pesticides in orchards, which have affected the production of
terrestrial invertebrates as food for juvenile salmon.
Frequency of Occurrence. Data on the number of stomachs containing one or more items
of a particular taxon (essentially a measure of presence or absence) allows an assessment
of the homogeneity of feeding habits in a population (Hyslop 1980). Baetid nymphs and
chironomids of all life stages were present in over 50% of the ﬁsh stomachs examined,
which conﬁrms their general importance expressed in terms of relative abundance. Other
items, such as nematodes and arachnids, were found in low abundances but were present
in over 10% of the stomachs, indicating that these were relatively rare but widely and
consistently available.
Prey Selection. Previous studies have shown prey selection by salmon to be based on size,
abundance, and visibility of prey (Higgs et al. 1995). Selection of prey by juvenile chinook
salmon in the Sacramento River, based on linear selection indices, showed that relatively
small or possibly undetectable organisms, such as crustaceans (Cladocera, Eucopepoda),
collembolans, and chironomid larvae, were avoided. Positively selected items included
chironomid pupae and adults, which are seasonally very abundant, and adult Baetidae,

which are relatively large as well as abundant. Such factors as surface movements during

50
adult emergence or conspicuous eyespots, in addition to size, may contribute to the

greater visibility and, hence, positive selection of these items. The presence of brightly
colored parasites may also increase visibility of prey organisms. For example, bright
orange larval hydracarinids are parasites on chironomid pupae, and usually crawl or fall
out of the pupal exuvia as the adult chironomid is emerging (Pennak 1978). In my study,
it appeared that chironomid pupae or emergent adults found in salmon stomachs had many
more of these parasites than the chironomids typically seen in the drift, although a formal
tabulation of these data was not attempted.

It should be noted from the above discussion that food selection indices are not
equivalent to an assessment of absolute preferences for certain taxa, but instead reﬂect a
combination of factors including preference, prior experience, prey detectability, and prey
availability. A limitation of all selection indices is that they cannot elucidate the underlying
mechanism(s) of the observed patterns of selection.

Fish Size, Gape Size, and Prey Size RglgtionshiLS. The gape size of juvenile chinook
salmon in the Sacramento River increased predictably with ﬁsh length (Figure 3). Larger
gapes would allow the salmon to potentially exploit a greater range of prey sizes as they
grow. Inclusion of increasingly larger (more proﬁtable) prey in the diet would, in turn, be
advantageous for future growth. Field-Dodgson (1988) reported that the mean mouth
breadth of emergent chinook ﬁy (2.2 mm gape; ranging from 2.0 to 2.5 mm; mean ﬁsh
size 33.5 mm FL) corresponded to the maximum HCW of Deleatidium nymphs found in
the gut, which suggests that juvenile chinook salmon will consume the largest prey
possible. In the Sacramento River, larger salmon ate larger prey on average, but the

maximum invertebrate prey sizes in the diet fell considerably short of ﬁsh gape potential,

51
as was observed in Atlantic salmon from Catamaran Brook, New Brunswick (Keeley and

Grant 1997). Wainwright and Richard (1995) proposed that pharyngeal gape rather than
oral gape may set the limit on the size of prey consumed, which would translate into
maximum prey being about halfthe size of mouth gape. However, as the authors noted,
distortion of prey during swallowing complicates that hypothesis.

The weak increase of mean and minimum invertebrate prey size with ﬁsh size
indicates that all juvenile salmon in the upper Sacramento River included small prey items
in their diet to meet energy requirements, despite the ability of larger ﬁsh to consume
larger prey. Essentially, invertebrate prey size was determined more by availability in the
environment than by morphological constraints. However, the ability of juvenile salmon
to capture ﬁsh larvae also appeared to increase as the salmon grew in size, but the actual
availability of such prey would be only indirectly measurable by drift methods. The
relationship between salmon size and piscivory is more complicated, as ﬁsh prey are
evasive and must be located and pursued, whereas drift feeding is more a process of
maintaining a stationary position suitable for intercepting drifting prey. Therefore, with
regard to piscivory, gape constraints as well as efficiency in pursuit or handling of ﬁsh
prey could explain the relationship between predator size and the number of prey
consumed.

Feedmg' Intensity and Stomgch F ullne_ss. Total number of items in the stomachs of
Sacramento River juvenile chinook were variable depending on the size of the ﬁsh and the
month of capture (Table 7). Nevertheless, due to prey size differences, the total number
of items offers little insight into the energy value of the items consumed (Hyslop 1980).

The stomach ﬁillness index provides a better indicator of energy intake. The mean SFI of

52
juvenile salmon from the Sacramento River was 2.3% of wet weight (iSD = 1.6; n =

189), comparable to other reports for juvenile chinook salmon feeding in the wild
(Brodeur 1992, Sagar and Glova 1988). Brodeur (1992) estimated a geometric mean
stomach ﬁtllness of 1.34% (n = 734) for juvenile chinook off the coast of Washington and
Oregon; the range of estimates, 0 to 8%, was very similar to the range in my study (0 to

8.4%). Sagar and Glova (1988) found mean percent dry weight of food per ﬁsh ﬂuctuated

I“

between 1.3 and 3.4% for juvenile chinook over one 24-hour period in the Rakaia River,
New Zealand. My estimate does not, however, reﬂect maximum feeding intensity, since
most samples were taken in the morning and early afternoon, whereas my 24-hour study
indicated maximum feeding occurred near dusk (see discussion below).
Drift grid Feedng over the Diel Cycle. The diel series, though unreplicated, offered some
insight into the daily patterns of feeding intensity of juvenile salmon in the Sacramento
River. The stomach fullness index was highest near dusk with a lesser peak before dawn,
whereas by number, feeding was highest near dusk. The result accords with that of Kolok
and Rondorf (1987), who determined that the greatest caloric intake by chinook salmon in
the Columbia River occurred between 1300 and 2100 hours. I estimated 24-hour
consumption by juvenile chinook salmon to be approximately 12.5% of wet body weight,
comparable to the 8.3% of dry weight estimated by Sagar and Glova (1988), and
corresponding to roughly half the maximum consumption for juvenile salmonids (about 17
to 20%) (Beauchamp et al. 1988). This estimate is reasonable for wild ﬁsh, which would
be expected to feed at less than maximum rates.

The differences in size and taxonomic composition of the Sacramento River drift

between day and night is a phenomenon widely observed in other river systems (Waters

 

53
1972, Brittain and Eikeland 1988) which results in changes in prey availability over the

diel period (Bowles and Short 1988, Sagar and Glova 1988). Decreased abundance of
baetid mayﬂies and increased abundance of small terrestrial organisms such as
collembolans during the daytime explains the difference in nocturnal and diurnal size
distribution in the Sacramento River drift. Though larger organisms were rare in the drift
during the daytime when juvenile chinook fed the most, larger prey were consumed in
numbers disproportionate to their presence in the drift.

Possible sources of error. Several factors may have affected drift and diet estimates.
These include processing error in drift samples, backwash in drift samples, sampling
constraints at night, and uneven representation of habitat types.

Drift density estimates were affected by the inclusion of a higher number of
microcrustaceans during January-March as compared with July-August and April-June,
which may be due either to differences in processing or environmental conditions, or both.

The importance of microcrustaceans to the diet was not detected in my study, but has
been shown for juvenile chinook in reservoir habitats of the Columbia River (Rondorf et
al. 1990). Based on the feeding habits of juvenile chinook in the Sacramento River
determined in my study, a more ecologically relevant quantiﬁcation of food availability
than that presented here would incorporate only taxa that the ﬁsh actually consume into
estimates of drift density.

The long duration of drift samples (15 min to over an hour) resulted in clogging of
the net mesh by sediment and algae, which produced some backwash. Though the
sampling interval was shortened in 1996 to improve this problem, inevitably some drift

densities were underestimated due to a loss of organisms in backwash.

 

54

Due to various constraints, nighttime drift and diet samples were taken on only one
occasion. Because the taxonomic composition of the drift and diet is not constant over
the diel cycle, my estimates of daytime feeding and drift should not be assumed to
represent nighttime feeding and drift. However, diet samples during the 24-hour study
showed that while nighttime feeding may not be important for juvenile chinook, feeding

during late afternoon and early evening was signiﬁcant. Depending on the objective at

”E

hand, future studies of chinook food and feeding in the Sacramento River could save time

and effort by conﬁning sampling of drift and diet to the period of highest feeding intensity.

My drift and diet results may not be representative of all riverine habitats occupied
by juvenile salmon in the Sacramento River. All drift samples were taken at gravel bar
sites rather than a mix of gravel bar and thalweg (edge) sites, which provide more cutbank
habitat and therefore a greater immediate contribution of terrestrial organisms. Fish were
also captured mostly at gravel bar sites, so the feeding habits in more protected habitats
are not described. Schaﬂter et al. (1982), however, found no difference in the diet of

juvenile chinook between edge and gravel bar sites.

Condition of Juvenile Chinook
Length-Weight Relationship. The length-weight relationship has been used both to
compare condition of ﬁsh samples or populations (Cone 1989) and to represent changes in
body form over ontogeny (Safran 1992). In discussing the length-weight relationship,
many authors implicitly assume that the line is unidirectional, with growth in weight
proceeding from the smallest length (LeCren 1951, Ricker 1979, Wootton 1992). Under

this assumption, breaks in the regression are interpreted as different “growth stanzas”

55
(Wootton 1992). Though this may be true for growth followed in individual ﬁsh,

designation of instantaneous samples from a ﬁsh population as representing an “isometric”
(b = 3.0) or “allometric” (b at 3.0) growth form (e.g., see Cone’s response in Springer et
al. 1990) may be inaccurate. For example, the length-weight relationship of juvenile
chinook salmon captured from the Sacramento River suggests that growth for 30 to 130
mm sizes is allometric (b = 3.49), with ﬁsh becoming increasingly heavy for their length
(Figure 7). Results from my growth experiments show, however, that the slope of the
length-weight regression is strongly inﬂuenced by negative as well as positive growth,
resulting in unusually high slopes in treatment groups fasted for two weeks, due to the
more severe effects of starvation on smaller ﬁsh (Appendix D). This unexpected result is
presumably due to the higher metabolic rate of smaller juvenile salmon. Godinho (1997)
recognized that variations in the slope of the length-weight relationship for Triportheus
guentheri could occur ﬁom individuals of different lengths changing their weight at
different rates or directions. Therefore, the slope or the intercept of the length-weight
relationship are meaningless as indices of condition when used alone (Bolger and Connolly
1989), and instead the entire length-weight relationship should be used (Cone 1989).
Authors should take care to report both slopes and intercepts to allow comparisons
between populations.

Due to its ambiguous nature, the slope of the length-weight relationship for
instantaneous samples is also questionable for understanding growth form, deﬁned by
Cone (1989) as the rate at which weight increases in relation to length. Inspection of
length-weight relationships within size classes appeared to indicate that growth in

Sacramento River juvenile salmon shifted ﬁom positive allometry to isometry or negative

56
allometry at smolt sizes (50 to 90 mm). Although this partly agrees with the general

notion of a change in body shape accompanying smoltiﬁcation (Hoar 1976), which
includes a longer caudal peduncle and narrower body (Beeman et al. 1994), there is little
reason why smaller ﬁsh, in the presence of adequate energy, would emphasize energy
storage or grth in weight (b > 3) rather than grth in length Length grth would
lead to a greater competitive advantage such as an increasing range of prey types
consumed, as I have shown in the discussion on predator-prey size relationships.

Because length-weight regression parameters for juvenile chinook salmon are not
commonly reported in the literature, juvenile salmon of known rearing history from my
growth experiments provide a useﬁll standard of comparison. Figure 9 shows that ﬁeld-
caught salmon were heavier at all lengths tlmn juvenile salmon that were fasted for two
weeks, but generally lighter than juveniles than fed a hatchery diet. The fact that wild
juvenile chinook were heavier—and therefore in better condition—than fasted ﬁsh of
similar length indicates that environmental conditions were conducive to feeding. Such a
result corroborates my ﬁndings on stomach fullness, which suggest that juvenile chinook,
while not feeding at maximum rates, were maintaining an adequate level of energy intake.
Experimental salmon fed low and high rations were in better condition than juvenile
chinook captured from the Sacramento River, however, this may be explained by
diﬂerences both in quality of food and in energy demands between experimental and ﬁeld-
caught ﬁsh.
ﬂigth-Dry Weight Relatimhip. The length-dry weight or length-percent water
relationships may be more useful than wet weight for representing true energy differences

among sizes. Percent dry weight or percent water can be used to estimate seasonal

57
changes in energy density of ﬁsh with reasonable conﬁdence (Hartman and Brandt 1995,

Jonas et a1 1996), due to the fact that changes in dry weight primarily reﬂect changes in
fat or protein, the most important energy stores in ﬁshes (Hayes and Taylor 1994). Given
the relationship between dry weight or percent water and energy, increase in dry Weight in
juvenile salmon over the range of lengths indicates that larger ﬁsh had greater proportions
of energy and lower proportions of body water, and hence greater energy density, than
smaller ﬁsh. Similar patterns were seen for condition factor (see below), in which larger
ﬁsh had higher condition factors than smaller ﬁsh. If percent water is considered
indicative of energy status, then indeed larger ﬁsh are in better condition than smaller ﬁsh.
What adaptive signiﬁcance could length-related differences in body composition
hold? One possible explanation has to do with predation risk or competitive interactions
(Gardiner and Geddes 1980). Salmon ﬁ'y are more vulnerable to predation due to small
size and undeveloped swimming ability, however use of a high proportion of water to
increase overall bulk when food is limited may make the difference between survival and
predation by gape-limited piscivores. Bulkier ﬁsh may also be more likely to “win” in
competitive interactions. As the sahnon grow to smolt sizes, swimming ability develops,
and survival of predation by avian and other non-gape-limited predators depends more
upon a quick escape than upon bulk. Therefore a streamlined body (without the water
bulk) associated with smoltiﬁcation and preparation for ocean existence has its advantages
in freshwater rearing. Differences in body composition among sizes may be associated
with the different risks of being large versus small.
Condition Factor (K). The length-weight relationship of Sacramento River juvenile

chinook salmon dictated that the range of observed condition factors was not constant

58

across lengths. However, as was discussed previously, a slope greater than 3.0 in the
length-weight relationship of a sample does not prove that the assumption of isometric
growth, considered necessary for proper use of K (LeCren 1951, Cone 1989), has been
violated. The greater limitation of the condition factor is that it does not allow
visualization of true differences in weight at length between samples or populations. My
estimates of condition factor for juvenile chinook salmon in the mainstem Sacramento
River were similar to mean condition factor reported for juvenile chinook in other systems
(Carl 1984, Hard 1986, Field-Dodgson 1988) and in intermittent tributaries of the
Sacramento River (Moore 1997).

Condition factors are often used to estimate the energy status of ﬁsh, and attempts
have been made to relate condition factor to percent fat, energy density, and other
components of proximate analysis, with varying results (Caulton and Bursell 1977,
Weatherley and Gill 1983, Herbinger and Friars 1991, Salam and Davies 1994, Jonas et
al. 1996). Condition factors in juvenile ﬁshes may be related to proximate components
because all factors vary with ﬁsh size. However, an extremely low (size-speciﬁc)
condition factor does reﬂect ecologically relevant differences in energy status. In my
study, experimental juveniles that were fasted for two weeks had extremely low condition
factors, were lethargic, and appeared to be swimming more slowly than ﬁsh in fed
treatment groups (Appendix D). The ability of the fasted ﬁsh to escape predation and to
locate and capture prey would have been very limited upon release into the demanding
river environment. However, with increased food availability and in the absence of high
energy demand for activity, fasted juvenile salmonids can recover and compensate with

rapid grth and increased condition (Weatherley and Gill 1981; see also Appendix D).

59
General Linea; Models Analysis of Spatial and Temmral Patterns. The patterns of salmon

length, weight, condition factor and percent water estimated by the least squares means
(GLM analysis) suggests that environmental conditions were conducive to ﬁsh growth and
increased condition over time and as the ﬁsh migrated downstream. However, due to the
unequal representation of size classes in samples from different months or sites, the GLM
analysis has little value for indicating true differences in weight or energy content at
length--which is what the ideal analysis of condition should do. That at least some ﬁsh
grew in length over time is obvious ﬁom the capture of larger ﬁsh as the 1996 ﬁeld season
progressed. However, the continual inﬂux of newly emerged sahnon, accompanied by the
outmigration of larger ﬁsh, would cause underestimation of the rate of growth based on
least squares means. Though the GLM did not work as an accurate estimate of growth
rate or a comparison of true size-speciﬁc condition across time and sites in my study, it
may adequately describe mean differences over time in situations not complicated by
protracted, overlapping emergence times of several stocks.

Analysis of Covariance of Length-Weight Relationships. Use of analysis of covariance to
compare sahnon body weight- or percent water-at-length between months and sites
conﬁrmed some patterns from the GLM analysis and reﬁtted others. Weight for a given
length (within the 50 to 90 mm size range) progressively increased from February through
June 1996 (Table 12), which may reﬂect a progressive increase in temperature and food
availability during that period. However, percent water at a given length was not
consistent across the three months studied, April-June 1996. The highest intercept
estimate of body water (which would indicate lowest energy density, or poorest condition)

was estimated for April, and the lowest was estimated for May. Table 13 reveals that

60

weight or percent water for a given length indicated no clear spatial trends in condition.
Though the mainstem is generally cooler in upstream sites with greater proximity to
Keswick Dam releases, the availability of warmer tributaries and backwaters for rearing
along the entire length of the study area may explain the variable estimates in weight-at-
Iength ﬁom upstream to downstream.

Possible Sources of Error. Fall-run size criteria based on a widely used growth model
(Fisher 1992, Greene 1992) dictated which ﬁsh were handled from beach seining sites in
April and May, whereas at other times, randomly selected ﬁsh of all sizes were measured
live to determine condition. Therefore the estimates of mean length and weight of
captured sahnon may suffer from nonrandom size selection during those months. Fish of
all sizes were represented in data from the rotary screw traps (RM 278) and Red Bluff

Research Pumping Plant (RM 242).

SUMMARY AND CONCLUSIONS

Is the upper Sacramento River a suitable rearing environment for juvenile chinook
salmon? Brodeur et al. (1992) posed the argument that “food limitation would more likely
manifest itselfin poor growth or condition of individual ﬁsh rather than mass starvation of
a cohort.” Only under uniformly severe food limitation or suboptimal temperatures would
all ﬁsh in a system manifest poor growth and condition. Therefore, based on my results
ﬁom 1995 and 1996, the habitat in the upper river was suitable for salmon rearing.

The numeric abundance and taxonomic composition of the food resources in the
Sacramento River were comparable to other river systems (Becker 1973, Bowles and
Short 1988, Sagar and Glova 1987, Sagar and Glova 1988), and the sahnon appeared to
have access to abundant foods, although the largest, most preferred sizes of invertebrates
were not always available. The lack of diet data for July-August 1995 and January-March
1996 precluded comparison of feeding habits between seasons. Since most fall-run ﬁ'y
leave the upper river and arrive at the Delta by April during wet years, food availability
during April-June may increase compared with January-March due to environmental
conditions (increased temperature and water clarity) as well as a reduction in juvenile
salmon density (intraspeciﬁc competition). Thus the juveniles that remained during April-
June 1996 may have beneﬁted ﬁom both these factors, and the observed feeding habits

reﬂect these favorable conditions.

61

62

Length-weight relationships, condition factors, and water content of juvenile
chinook sahnon in the upper Sacramento River indicate that environmental conditions
should be generally favorable for good growth and robustness during wet years similar to
1995 and 1996. Wet years are characterized by an increase in habitat and greater'lateral
habitat complexity in the form of backwaters and intermittently-ﬂowing tributaries, as
compared with dry years (Schlosser 1991 ). Though it is not possible to separate ﬁsh
based on the primary habitat(s) responsible for grth (hatchery, tributary, mainstem or a
combination of these), the sahnon in general were in better condition (heavier at all
lengths) than sahnon fasted for two weeks under experimental conditions. My growth
experiments also show that the smallest sizes could be more aﬁ‘ected by food limitation.
Still, some juveniles reared in the Sacramento River system in 1996, speciﬁcally in the
tributaries, grew at rates near the published maximum for juvenile chinook in the wild
(Moore 1997). An interesting question, unanswerable with my data, concerns the length-
weight relationships and condition factors of juvenile sahnon during dry and critically dry
years. Dry years would be characterized by a higher initial density of ﬁsh due to lower
dispersal of ﬂy to the Delta, and by lower habitat area. Further studies could focus on
comparing the growth and condition of ﬁsh in different water-years in different parts of
the Sacramento River (upper, middle, lower, and the Delta).

Based on my results, juvenile chinook sahnon would be relatively competent, in an
energetic sense, to withstand some of the natural stressors they would encounter on their
way out of the upper river. Energetic status only partially determines a ﬁsh’s ability to
survive, however, particularly when water development in the river is considered.

Mortality risk may be higher for smaller salmon, even those in relatively good condition

63

for their size, due to greater vulnerability to piscivorous ﬁsh waiting below dams (Garcia
1989) and lower screening efﬁciency at water diversions (Clark and Strong 1991). Other
factors, such as elevated water temperatures in the lower river, Delta water diversions,
and entrainment in Delta pumps may reduce sahnon survival regardless of condition
(Kjelson and Brandes 1989).

I recommend continued monitoring of the condition of juvenile Sacramento River
salmon using length-weight relationships, condition factors, and percent water estimates,
which provide a relatively easy way to track changes in the health of the population and
the habitat suitability of the upper river system. These measures could be readily
incorporated into existing sampling programs, such as regular seining by the US. Fish and
Wildlife Service. Detailed calorie and fat analyses would be desirable for determining the
relationship between energy content and the less labor-intensive measures of condition,
which would conﬁrm the usefulness of these indices for tracking seasonal and annual

changes in energy status of Sacramento River chinook salmon.

APPENDICES

APPENDIX A

APPENDIX A

Relative Abundances of Drift Taxa July-August 1995 and January-March 1996

Table 14. Total numbers and relative percentages of organisms captured in the drift from
the Sacramento River. July-August 1995 (n = 84 samples), and late January-March 1996
(n = 74 samples) (“-” indicates item not present).

 

 

 

Taxon Life July-Aug. July-Aug. Jan-March Jan-March

Stage 1 995 1995 1996 1996

__ _ _ __ ________ Number Percent Number Percent

Class Insecta
Order Diptera
Chironomidae Adult 1432 l 1.68 210 2. 70
Simuliidae Adult 3 1 0.25 - —
Dolichopodidae Adult - - 1 0.01
Empididae Adult 1 0.01 - -
Ephydridae Adult 1 0.01 3 0.04
Psychodidae Adult 1 0.01 2 0.03
Cecidomyiidae Adult 12 0.10 l 2 0. 1 5
Synneuridae Adult 1 0.01 - -
Phoridae Adult 1 0.01 - -
Unidentiﬁed Adult 1 0 0.08 3 0.04
Chironomidae Pupa 740 6.03 72 0.93
Ceratopogonidae Pupa 6 0.05 - -
Simuliidae Pupa 3 0.02 - -
Unidentiﬁed Pupa 3 0.02 - -
Chironomidae Larva 6307 51 .43 1268 16.32
Ceratopogonidae Larva 1 0.01 6 0.08
Simuliidae Larva 700 5.71 298 3.84
Dolichopodidae Larva l 0.01 l 0.01
Empididae Larva 3 0.02 7 0.09
Ephydridae Larva - - l 0.01
Tipulidae Larva 8 0.07 45 0.58
Dixidae Larva - - l 0.01
Muscidae Larva 2 0.02 1 0.01
Stratiomyidae Larva l 0.01 6 0.08
Culicidae Larva - - l 0.01
Chaoboridae Larva - - 1 0.01
Unidentiﬁed Larva l 0.01 5 0.06
TOTAL DIPTERA All 9266 75.55 1944 25.03

Stages

 

64

Table 14 (cont’d)

65

 

 

Taxon Life July-Aug. July-Aug. Jan-March Jan-March
Stage 1995 1995 1996 1996
Number Percent Number Percent
Order Coleoptera
Carabidae Adult 2 0.02 - -
Dytiscidae Adult 1 0.01 21 0.27
Staphylinidae Adult 1 0.01 4 0.05
Hydraenidae Adult - - 1 0.01
Ptiliidae Adult 1 0.01 - -
Anthicidae Adult 1 0.01 - -
Circulionidae Adult 1 0.01 - -
Cerambycidae Adult - - 1 0.01
Scolytidae Adult - - 1 0.01
Unidentiﬁed Adult 2 0.02 5 0.06
Gyrinidae Larva 3 0.02 - -
Dytiscidae Larva 7 0.06 6 0.08
Staphylinidae Larva 1 0.01 3 0.04
Hydrophilidae Larva 1 0.01 3 0.04
Ptiliidae Larva l 0.01 - -
Elrnidae Larva 2 0.02 3 0.04
Unidentiﬁed Larva 2 0.02 3 0.04
TOTAL All 26 0.24 5 1 0.66
COLEOPTERA Stages
Order
Ephemeroptera
Caenidae Adult 2 0.02 - -
Baetidae Adult 135 l. 10 16 0.21
Unidentiﬁed Adult 7 0.06 - -
Oligoneuriidae Nymph l 0.01 - -
Tricorythidae Nymph 10 0.08 87 1.12
Caenidae Nymph - - l 5 0. l9
Ephemerellidae Nymph 6 0.05 38 0.49
Heptageniidae Nymph 1 2 0. 10 19 0.24
Baetidae Nymph 1829 14.91 690 8.88
Siphlonuridae Nymph 1 0.01 4 0.05
Leptophlebiidae Nymph - - l 0.01
Unidentiﬁed Nymph 8 0.07 1 0.01
TOTAL All 2011 16.40 871 11.21
Ephemeroptera Stages

 

Table 14 (cont’d)

66

 

 

Taxon Life July-Aug. July-Aug. Jan-March Jan-March
Stage 1995 1995 1996 1996
Number Percent Number Percent
Order Hemiptera
Corixidae Adult 2 0.02 2 0.03
Naucoridae Adult 1 0.01 - -
Berytidae Adult 1 0.01 - -
Lygaeidae Adult 1 0.01 - -
Miridae Adult 1 0.01 - -
Tingidae Adult - - 1 0.01
Anthocoridae Adult 1 0.01 - -
Unidentiﬁed Adult 2 0.02 - -
Corixidae Nymph 80 0.65 1 0.01
Berytidae Nymph - - l 0.01
Unidentiﬁed Nymph 6 0.05 7 0.09
TOTAL All 95 0.77 12 0.15
HEMIPTERA Stages
Order Homoptera
Cicadellidae Adult 3 0.02 2 0.03
Psyllidae Adult 7 0.06 - -
Aphididae Adult 42 0.34 8 0. 10
Coccoidea Adult 1 0.01 23 0.30
Unidentiﬁed Adult 1 0.01 1 0.01
Cercopidae Nymph - - 1 0.01
Cicadellidae Nymph 2 0.02 7 0.09
Psyllidae Nymph 3 0.02 3 0.04
Aphididae Nymph 14 0.1 1 6 0.08
Fulgoroidea Nymph - - 3 l 0.40
Unidentiﬁed Nymph 2 0.02 l 3 0. 17
TOTAL All 75 0.61 95 1.22
HOMOPTERA Stages
Order Hymenoptera
Forrnicidae Adult 1 0 0.08 10 0. 13
Mymaridae Adult - - 2 0.03
Trichogrammatidae Adult 1 0.01 - -
Eucharitidae Adult - - 1 0.01
Unidentiﬁed Adult 1 5 0. 12 10 0.13
TOTAL 26 0.21 23 0.30

HYMENOPTERA

 

Table 14(cont’d)

67

 

 

Taxon Life July-Aug. July-Aug. Jan-March Jan-March
Stage 1995 1995 1996 1996
Number Percent Number Percent .
Order Lepidoptera
Pyralidae Larva - - 1 0.01
Nepticulidae Larva - - 3 0.04
Unidentiﬁed Larva 1 0.0 l - -
TOTAL 1 0.01 4 0.05
LEPIDOPTERA
Order Odonata
Coenegrionidae Nymph - - 7 0.09
Libellulidae Nymph - - l 0.01
TOTAL - - 8 0.10
ODONATA
Order Plecoptera
Taeniopterygidae Nymph - - 8 0. 10
Nemouridae Nymph - - 21 0.27
Capniidae Nymph - - 63 3 8. 1 5
Perlidae Nymph - - 2 0.03
Chloroperlidae Nymph 1 0.01 - -
Perlodidae Nymph 3 0.02 85 1 .09
Unidentiﬁed Nymph 6 0.05 1 1 0. 14
TOTAL 10 0.08 760 9.78
PLECOPTERA
Order Thysanoptera
Phlaeothripidae Adult 2 0.02 6 0.08
Aeolothripidae Adult 2 0.02 - -
Thripidae Adult 6 0.05 1 0 0. 13
Unidentiﬁed Pupa - - 3 0.04
TOTAL All 10 0.08 19 0.24
THYSANOPTERA Stages
Order Trichoptera
Hydroptilidae Adult 4 0.03 7 0.09
Glossosomatidae Adult 1 0.01 3 0.04
Psychomyiidae Adult 5 0.04 - -
Brachycentridae Adult 1 0.01 l 0.01
Unidentiﬁed Adult - - 2 0.03

 

Table 14 (cont’d)

68

 

 

Taxon Life July-Aug. July-Aug. Jan-March Jan-March
Stage 1995 1995 1996 1996
Number Percent Number Percent -
Glossosomatidae Pupa - - 1 0.01
Hydropsychidae Larva 47 0.38 89 1 . l 5
Hydroptilidae Larva 2 0.02 5 0.06
Leptoceridae Larva - - 3 0.04
Glossosomatidae Larva - - 63 0.81
Philopotamidae Larva 1 0.01 - -
Psychomyiidae Larva 3 0.02 36 0.46
Brachycentridae Larva 1 0.01 20 0.26
Odontoceridae Larva - - 1 0.0 l
Limnephilidae Larva - - 4 0.05
Unidentiﬁed Larva 5 0.04 6 0.08
TOTAL All 70 0.57 241 3.10
TRICHOPTERA Stages
Order Orthoptera
Tridactylidae Nymph l 0.01 - -
Unidentiﬁed Nymph 2 0.02 - -
Order Collembola All 6 0.05 84 1.08
Stages
Order Entotrophi
Japygidae - - 4 0.05
Unidentiﬁed All - - l 0.01
Stages
Other Unidentiﬁed Adult - - l 0.01
Insects
Larva/ - - 7 0.09
Nymph
Other Items
Petromyzontidae Larva - - 1 0.01
Teleostei Larva 146 1 .19 29 0.37
Teleostei Egg - - 2 0.03
Hydroida 4 0.03 74 0.95
Nematoda 14 0.11 30 0.39
Nematomorpha - - 4 0.05
Oligochaeta 390 3 . 1 8 448 5.77

 

Table I4 (cont’d)

69

 

 

Taxon Life July-Aug. J uly-Aug. Jan-March Jan-March
Stage 1995 1995 1996 1996
Number Percent Number Percent -
Polychaeta - - 2 0.03
Anostraca - - 1 0.01
Cladocera
Bosminidae - — 13 5 1 .74
Daphniidae 5 0.04 75 0.97
Macrothricidae - - 3 0.04
Chydorinae 4 0.03 - -
Unidentiﬁed 10 0.08 97 1.25
Cladocera
Eucopepoda
Calanoida 4 0.03 949 12.22
Cyclopoida 32 0.26 1449 18.65
Harpacticoida - - l 0.01
Ostracoda 6 0.05 70 0.90
Isopoda 1 0.01 - -
Decapoda - - l 0.01
Amphipoda 2 0.02 14 0. l 8
Arachnida 19 0.15 14 0.17
Hydracarina 28 0.23 233 3.00
Gastropoda - - 9 0. 12
Unidentiﬁed - - 2 0.02
TOTAL OTHER 665 5.42 3643 46.90
ITEMS
GRAND TOTAL 12264 100 7768 100

 

APPENDIX B

APPENDIX B

Relative Abundances of Drift and Diet Taxa April-June 1996

Table 15. Relative abundance of items in the drift (n = 113 samples) and diet of juvenile
chinook (n = 189 ﬁsh) in the Sacramento River for the period April-June 1996. Two ﬁsh
had empty stomachs (“-”

indicates an item not present).

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number” Percent Number Percent

Class Insecta

Order Diptera

Tipulidae Adult - - 9 0. 1 3
Simuliidae Adult 24 0.21 44 0.62
Psychodidae Adult 1 0.01 - -
Ceratopogonidae Adult 1 0.01 - -
Chironomidae Adult 1 305 l 1 .28 2059 28.95
Empididae Adult 2 0.02 43 0.60
Ephydridae Adult 1 1 0. 10 24 0.34
Muscidae Adult - - 1 0.01
Cecidomyiidae Adult 23 0.20 3 0.04
Unidentiﬁed Adult 27 0.23 20 0.28
Tipulidae Pupa 1 0.01 3 0.04
Simuliidae Pupa - - l 5 0.21
Ceratopogonidae Pupa 2 0.02 4 0.06
Chironomidae Pupa 241 2.08 923 12.98
Empididae Pupa - - 4 0.06
Unidentiﬁed Pupa - - 2 0.03
Tipulidae Larva 5 0.04 7 0.10
Blephariceridae Larva 2 0.02 - -
Simuliidae Larva 440 3.80 161 2.26
Tanyderidae Larva - - 1 0.01
Ceratopogonidae Larva 11 0.10 1 0.01
Chironomidae Larva 4294 37.12 1440 20.25
Dolichopodidae Larva - - 1 0.01
Empididae Larva 2 0.02 - -
Muscidae Larva - - 6 0.08
Stratiomyidae Larva l 0.01 - -
TOTAL DIPTERA All 6393 55.27 4771 67.08

Stages

 

70

Table 15 (cont’d)

71

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number Percent Number Percent
Order Coleoptera
Carabidae Adult 3 0.03 - -
Dytiscidae Adult 5 0.04 - -
Staphylinidae Adult 24 0.21 1 0 0. 14
Hydrophilidae Adult 4 0.03 - —
Hydraenidae Adult 2 0.02 - -
Ptiliidae Adult 1 0.01 - -
Scarabaeidae Adult 4 0.03 1 0.01
Melandryidae Adult 1 0.01 - -
Lyctidae Adult - - l 0.01
Unidentiﬁed Adult 1 2 0.10 4 0.06
Gyrinidae Larva l 0.01 - -
Dytiscidae Larva 14 0.12 5 0.07
Hydrophilidae Larva - - 3 0.04
Ptiliidae Larva 2 0.02 - -
Elmidae Larva 1 0.01 - -
Unidentiﬁed Larva 1 0.01 3 0.04
TOTAL All 75 0.65 27 0.38
COLEOPTERA Stages
Order
Ephemeroptera
Tricorythidae Adult 9 0.08 l 0.01
Caenidae Adult 20 0. 1 7 - -
Ephemerellidae Adult - - 8 0.1 1
Baetidae Adult 133 1.15 171 2.40
Unidentiﬁed Adult 8 0.07 33 0.46
Tricorythidae Nymph 1 1 0. 10 9 0. 1 3
Caenidae Nymph 2 0.02 - -
Ephemerellidae Nymph 5 0.04 - -
Heptageniidae Nymph 12 0. 1 0 - -
Baetidae Nymph 1826 15.79 1006 14.15
Siphlonuridae Nymph 5 0.04 5 0.07
Unidentiﬁed Nymph 8 0.07 12 0.17
TOTAL All 2039 17.63 1245 17.51
Ephemergitera Stages

 

 

 

Table 15 (cont’d)

72

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number Percent Number Percent
Order Hemiptera
Corixidae Adult 2 0.02 3 0.04
Saldidae Adult 1 0.01 - -
Hebridae Adult 2 0.02 - -
Macroveliidae Adult 9 0.08 - -
Berytidae Adult 1 0.01 - -
Unidentiﬁed Adult 2 0.02 1 0.01
Corixidae Nymph 26 0.22 5 0.07
Saldidae Nymph 1 0.01 - -
Hebridae Nymph 1 0.01 - -
Cimicidae Nymph - - l 0.01
Unidentiﬁed Nymph 5 0.04 5 0.07
TOTAL All 50 0.43 15 0.21
HEMIPTERA Stages
Order Homoptera
Cicadellidae Adult 25 0.22 23 0.32
Psyllidae Adult 3 0.03 - -
Aphididae Adult 71 0.61 167 2.35
Phylloxeridae Adult 1 0.01 - -
Coccoidea Adult 1 8 0. 1 6 - -
Unidentiﬁed Adult 1 0.01 23 0.32
Cercopidae Nymph 1 0.01 - -
Cicadellidae Nymph 3 0.03 3 0.04
Aphididae Nymph 3 0.03 - -
Unidentiﬁed Nymph 1 3 0.1 1 6 0.08
TOTAL All 139 1.20 222 3.12
HOMOPTERA Stages
Order Hymenoptera
Apidae Adult 2 0.02 - -
Forrnicidae Adult 19 0.16 15 0.21
Mymaridae Adult 2 0.02 - -
Pteromalidae Adult 2 0.02 - -
Ichneumonidae Adult - - 2 0.03
Braconidae Adult 6 0.05 1 0.01
Eulophidae Adult - - 1 0.01
Torymidae Adult 2 0.02 - -
Unidentiﬁed Adult 73 0.63 79 l . 1 1

 

73

Table 15 (cont’d)

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number Percent Number Percent
TOTAL 106 0.92 98 1.38
HYMENOPTERA
Order Plecoptera
Taeniopterygidae Adult 1 0.01 - -
Perlodidae Adult 7 0.06 2 0.03
Unidentiﬁed Adult - - 3 0.04
Nemouridae Nymph 1 0.01 - -
Capniidae Nymph 2 0.02 l 0.01
Perlidae Nymph - - l 0.01
Chloroperlidae Nymph l 0.01 - -
Perlodidae Nymph 6 0.05 14 0.20
Unidentiﬁed Nymph 3 0.03 7 0. 10
TOTAL All 21 0.18 28 0.39
PLECOPTERA Stages
Order Thysanoptera
Phlaeothripidae Adult 1 0.01 - -
Aeolothripidae Adult 2 0.02 - -
Thripidae Adult 4 0.03 1 0.01
Unidentiﬁed Adult 1 0.01 - -
TOTAL 8 0.07 1 0.01
THYSANOPTERA
Order Trichoptera
Helicopsychidae Adult 1 0.01 - -
Hydropsychidae Adult 5 0.04 0 0.84
Hydroptilidae Adult 7 0.06 8 0.1 1
Leptoceridae Adult 2 0.02 1 0.01
Glossosomatidae Adult 1 0.01 8 0.1 l
Philopotamidae Adult - - 2 0.03
Psychomyiidae Adult 1 8 0.16 5 0.07
Brachycentridae Adult - - 1 0.01
Lepidostomatidae Adult 1 0.01 - -
Limnephilidae Adult 1 0.01 - -
Unidentiﬁed Adult 1 0.01 27 0.38
Hydropsychidae Pupa - - l 0.01
Psychomyiidae Pupa 1 0.01 -
Hydropsychidae Larva 48 0.41 l .98
Glossosomatidae Larva 17 0.15 0.3 5

 

Table 15 (cont’d)

74

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number Percent Number Percent
Philopotamidae Larva l 0.01 1 0.01
Polycentropodidae Larva l 0.01 - -
Psychomyiidae Larva 1 1 0.10 2 0.03
Brachycentridae Larva 1 1 0. 1 0 - -
Limnephilidae Larva 1 0.01 - -
Unidentiﬁed Larva 21 0. l 8 3 0.04
TOTAL All 149 1.29 285 4.01
TRICHOPTERA Stages
Order Collembola All 1 12 0.97 2 0.03
Stages
Order Derrnaptera l 0.01 - -
Order Lepidoptera
Unidentiﬁed Adult 3 0.03 - -
Pyralidae Larva 1 0.01 - -
Unidentiﬁed Larva - - 7 0. 1 0
TOTAL All 4 0.04 7 0.10
LEPIDOPTERA Stages
Order Odonata
Libellulidae Nymph 1 0.01 - -
Order Orthoptera
Gryllidae Nymph l 0.01 - -
Order Psocoptera Adult 24 0.21 - -
Other Unidentiﬁed Adult 3 0.03 2 0.03
Insects
Larva/ 2 0.02 7 0.10
Nymph
Egg 1 0.01 - -
mass
Other Items
Catostomidae Larva 77 0.67 37 0.52
Cottidae Larva 3 0.03 - -
Petromyzontidae Larva 5 0.04 - -

 

Table 15 (cont’d)

75

 

 

Taxon Life April-June April-June April-June April-June
Stage 1996 Drift 1996 Drift 1996 Diet 1996 Diet
Number Percent Number Percent
Teleostei Larva 403 3.48 170 2.39
Teleostei Eggs - - 10 0.14
Hydroida 105 0.91 1 0.01
Nematoda 55 0.48 100 1.41
Oligochaeta 618 5.34 13 0.18
Cladocera
Bosminidae 1 l 0. 1 0 - -
Daphniidae 5 0.04 - -
Macrothricidae 340 2.94 - -
Unidentiﬁed 147 1 .27 6 0.08
Eucopepoda
Calanoida 1 19 1 .03 - -
Cyclopoida 419 3 .62 - -
Harpacticoida 1 0.01 - -
Ostracoda 6 0.05 - -
Isopoda 2 0.02 1 0.01
Decapoda l 0.01 - -
Amphipoda 4 0.03 2 0.03
Arachnida 33 0.29 28 0.39
Hydracarina 82 0.71 4 0.06
Gastropoda 2 0.02 - -
Unidentiﬁed - - 30 0.42
TOTAL OTHER 2438 21.08 402 5.65
ITEMS
GRAND TOTAL 11567 100 7112 100

 

APPENDIX C

APPENDIX C

Mean Length, Weight, Condition, and Percent Water of
Sacramento River Juvenile Chinook Salmon

Table 16. Mean fork length ( +/-SE), mean weight (+/- SE), mean condition factor (K)
(+/-SE), and mean percent water (+/-SE) (where applicable) of young-of-year chinook
salmon captured at study sites on the Upper Sacramento River, July 1995 and February-
June 1996. Mean length, weight and condition reported for certain sites sampled in April
and May have been adjusted for delays in measurement 4-11 hrs postmortem. In cases
where the number of ﬁsh used for determination of percent water is different from the
number used for length, weight and condition calculations, the n for percent water is
reported in parentheses. Mean percent water was calculated based on the thawed weight
of ﬁsh taken just prior to drying.

 

 

Mean Mean
Month . n (n Fork Weight Mean K Percent
and Site Length
Year water) (m) (g) (1813) Water
-____ (i SE) (iSE) (iSE)
July Bidwell 17 83 (2) 6.9 (0.5) 1.2 (0.01)
1995
Sacramento 49 76 (2) 5.2 (0.3) 1.1 (0.01)
Bar
Below 3 55(1) 1.3 (0.1) 0.8 (0.01)
RBDD
Feb. Bidwell 27 48 (4) 1.75 (0.65) 0.89 (0.02)
1996
Woodson 27 53 (5) 2.65 (0.97) 0.84 (0.03)
Sacramento 41 45 (2) 1.23 (0.51) 0.83 (0.02)
Bar
Below 35 45 (2) 1.09 (0.34) 0.85 (0.02)
RBDD
Pizza Place 14 39 (2) 0.48 (0.09) 0.75 (0.03)
Bend 21 43 (5) 1.63 (1.12) 0.80 (0.03)
Anderson 35 38 (1) 0.43 (0.03) 0.77 (0.02)
Posse 33 39 (l) 0.50 (0.04) 0.78 (0.02)
March Bidwell 10 46 (2) 0.99 (0.13) 0.92 (0.03)
1996
Woodson 15 50 (2) 1.23 (0.13) 0.93 (0.02)
Sacramento 20 50 (2) 1.24 (0.16) 0.94 (0.02)
Bar
Pizza Place 4 53 (3) 1.38 (0.22) 0.91 (0.03)
Bend 9 48(4) 1.31 (0.35) 1.01 (0.04)
Balls Ferry 6 44 (3) 0.86 (0.17) 0.91 (0.05)

Mean

 

 

76

Table 16 (cont’d)

77

 

Mean

 

Mean Mean
Month . n (n Fork Weight Mean K Percent
and Site Length
water) (g) (iSE) Water
Year (mm) (+SE) +SE
_ (i SE) - (- )

April Bidwell 39 (38) 63 (2) 2.95 (0.30) 0.99 (0.02) 83.99 (0.65)
1996

Woodson 38 67 (2) 3.25 (0.23) 1.00 (0.02) 84.19 (0.23)

Sacramento 40 55 (2) 1.92 (0.21) 0.97 (0.02) 83.59 (0.36)

Bar

Below 20 (15) 72 (2) 4.05 (0.28) 1.05 (0.01) 81.01 (0.33)

RBDD

Red Bluff 189 55(1) 2.25 (0.15) 0.91 (0.01)

Research

Pumping

Plant

Above 70 64 (2) 3.61 (0.30) 1.02 (0.02)

RBDD West

Pizza Place 20 68 (2) 3.36 (0.31) 1.02 (0.01) 80.47 (0.33)

Bend 25 (14) 51 (3) 1.51 (0.24) 0.90 (0.02) 83.17 (0.74)

Battle 15 61 (3) 2.51 (0.31) 1.02 (0.03) 83.82 (0.39)

CAFG Screw 299 (143) 55 (1) 2.29 (0.13) 0.92 (0.01) 81.92 (0.25)

Traps
May Bidwell 39 (36) 73 (1) 4.50 (0.22) 1.14 (0.01) 80.02 (0.27)
1996

Woodson 20 69 (l) 3.66 (0.22) 1.10 (0.01) 80.87 (0.16)

Sacramento 27 (25) 65 (2) 3.15 (0.30) 1.08 (0.01) 81.34 (0.26)

Bar

Pizza Place 21 62 (2) 2.66 (0.32) 1.01 (0.01) 82.19 (0.65)

Battle 12 55 (2) 1.91 (0.26) 1.06 (0.02) 84.58 (0.29)

CAFG Screw 52 (36) 67 (2) 3.60 (0.33) 1.02 (0.01) 80.54 (0.33)

Traps
June Bidwell 9 74 (4) 5.08 (0.77) 1.17 (0.01) 81.36 (0.66)
1996

Woodson 30 (10) 72 (3) 5.03 (0.49) 1.17 (0.02) 82.39 (0.39)

Sacramento 32 (l 8) 80 (1) 6.33 (0.29) 1.20 (0.01) 79.04 (0.31)

Bar

CAFG Screw 25 (12) 84 (2) 6.81 (0.50) 1.10 (0.02) 78.85 (0.46)

Traps

 

APPENDIX D

APPENDIX D

Growth Experiments with Hatchery Salmon

Purpose

The interpretation of condition data depends on the species- and population-
speciﬁc determination of “good” vs. “poor” condition. I conducted two separate growth
trials to assist in making the appropriate statements regarding ﬁsh captured in the ﬁeld.
The purpose of the experiments was to deﬁne the relationship between condition and feed
level or ration, which could then be used to interpret condition data, and thus infer
energetic status, from juvenile chinook salmon of unkown feeding history captured in the
ﬁeld. The speciﬁc goal of the experiments was to determine how the length-weight
relationship, condition factor, and percent body water of juvenile chinook salmon varies

with ration under experimental conditions.

Methods
The experiments included one fasted treatment group and one or two fed
treatment groups. The ﬁrst experiment had two treatments in which the ﬁsh were fed, the
second experiment had only one fed treatment group. The experiments were conducted
at different times and also differed in the size range of ﬁsh used. Table 15 summarizes
treatment information for both experiments.
The experiments used Coleman National Fish Hatchery fall-run chinook held in

210-gal ﬁberglass circular tanks supplied with aerated well water. The temperature was

78

79
held constant at 16°C. Each treatment in Growth Experiments 1 and II began with

approximately 300 ﬁsh per tank. Random subsamples of 10 to 20 ﬁsh were removed
from each tank once per week, and the zero-ration treatments were halted before
signiﬁcant mortality could occur. BioMoist grower pellets were distributed to fed
treatment groups on a twelve-hour belt feeder (1.0 to 2.4 mm pellets were used,
depending on ﬁsh size). Initial rations were calculated based on the length-weight
relationship of the initial subsamples, and the rations were adjusted according to the
number of ﬁsh remaining in the tank after subsamples were taken.

Fish that were selected as part of a subsample were anaesthetized until dead with
a solution of tricaine methanesulfonate (MS-222). The length and weights were recorded
and used to calculate condition factor (K). The ﬁsh were frozen in a bag of water and
later thawed, their thawed lengths and weights were recorded, and the carcasses were
dried to a constant weight at 70°C. The dry weights were used to calculate percent body
water based on thawed wet weight.

Results

Experiment I. General linear models results showed a highly signiﬁcant day, treatment,
and day x treatment effect on the length, weight, condition factor K, and percent body
water of salmon used in Experiment I (P < 0.0001 for all variables). The three factors
accounted for a high portion of the variability in length, weight, condition, and percent
water in Experiment I, with coefﬁcients of determination (r2) of 0.78, 0.89. 0.89, and

0.90, respectively.

 

80

Table 17. Summary of treatment information for growth experiments with juvenile

chinook salmon.

 

 

EXPERIMENT l EXPERIMENT 11
Number of Treatments 3 2
Ration Level (percent body
weight fed per day)
Treatment 2 4 to 10% Not applicable
Treatment 1 2 to 5% 3 to 4%
Treatment 0 0% 0%
Initial number of ﬁsh per tank 300 300
Mean initial length of 44 68
experimental ﬁsh (mm)
Range (mm) (40 to 47) (58 to 75)
Duration growth was 55* 29
monitored (days)
Duration of zero ration 16 15
treatment (days)
Recovery growth of Treatment No* Yes
0 monitored?
Recovery ration for unfed Not applicable 3 to 7% (same ration in grams

treatments (% body weight fed
per day)

as Treatment 1 during the same
period)

 

*Fish from Treatment 0, Experiment I, were accidentally released into the river prior to
the intended recovery feeding period. Beyond day 16, only the growth of treatments 1

and 2 was monitored.

81
The treatment group fed the highest ration in Experiment 1 (Level 2) had a 22.5%

increase in length, growth rate of about 0.71 mm/d over the two-week period of interest.
The Level 2 group more than doubled their weight (143% increase) on average, growing
about 0.07 g/d over two weeks. Level 1 ﬁsh increased in length by 13.6% and in weight
by 79.4%, growing about 0.42 mm/d and 0.03 g/d. In contrast, juveniles of the zero
ration group (Level 0) did not grow in length (an average 1.81% decrease was detected),
and lost 0.01 g/d, or a total loss of 22.4% of their initial body weight over the duration of
the treatment.

The initial and post-treatment length-weight relationships for ﬁsh in Experiment I

are shown in Figure 11.

Experiment 11. General linear models analyses for Experiment 11 data revealed a less
marked effect of day on length, weight, condition factor K, and percent water (P < 0.02).
However, the treatment effect was highly signiﬁcant (p < 0.003), as was the interaction
term (P < 0.0001). Values of r2 showed that 26,45, 70, and 58% ofthe variation in
length, weight, condition, and percent water could be explained by the effects studied.
The percent water estimates on the ﬁrst day of Experiment 11. were found to be
signiﬁcantly different between the two treatments (p < 0.0003, r2 = 0.23), but length.
weight, and condition were not signiﬁcantly different on day one.

The ﬁsh used in Experiment 11 demonstrated trends in growth similar to those
seen in Experiment I ﬁsh. Level 1 ﬁsh increased in length by 12.4% overall, or about
0.55 mm/d, and gained 0.12 g/d, with an overall increase of 57.5 % in weight. Fish that
were fasted for two weeks (Level 0) lost 27.9% of their total body weight on average, or

about 0.06 g/day, by the end of the treatment. As in Experiment 1, ﬁsh of the zero ration

82
group appeared to show a modest decrease in length (3.38% decrease). During the two-

week recovery period, during which Level 0 ﬁsh were fed the same ration (in grams) as
Level 1 ﬁsh, the Level 0 group increased in length by 15% (0.71 mm/d) and doubled their
weight (102.9% increase, or 0.17 g/d). By the end of the recovery period, the zero ration
group had attained an average length and weight comparable to what Level 1 ﬁsh had
reached by the end of the ﬁrst phase of the experiment. While the Level 0 ﬁsh were
recovering, Level 1 ﬁsh grew in length by 14.6% (0.78 mm/d) and increased in weight by
43.8% (0.16 g/d).

The length-weight relationships for ﬁsh in Experiment 11 at the end of week 2 are

shown in Figure 12.

In Wet Weight (mg)

83

 

      
  

 

 

8
y = 2.7797x - 3.6485
R2 = 0.914
7.5 .
y = 2.9809x - 4.5156 F
2 -
7 % R — 0.9563 1
I
I
y = 2.2377x - 1.9314 !
R2 = 0.6096
6.5 ~
I 0 Level 2
l 1:: Level 1
A l
6 .. y = 3.5225x - 7.0403 = a Level 0
a R2 = 0.7416 1
a i o lnitial Sample
5.5 i i i i l r i t

 

3.65 3.7 3.75 3.8 3.85 3.9 3.95 4 4.05 4.1
In Fork Length (mm)

Figure 11. Initial and ﬁnal (after two weeks) In length-In weight relationships for
the three treatment groups in Growth Experiment 1. Level 0 = zero ration
treatment; Levels 1 and 2 were fed. One outlier was excluded from level 0.

 

In Wet Weight (mg)

84

    
    

9.5 » . - L ...... -
1
1:1
1 y = 2.8558x - 3.8373
9* R2=0.9202
l
8.5 1:1
‘ y = 3.5024x - 6.6759
~ R’=0.8891
8 l
l . ..
7.5 . iDLevel1
y = 4.0178x - 9.09 1‘ Leve' 0 i
R2 = 0.9389 lo Initial Sample}
7 l
l
6.5 . er , - .. , .
3.9 4 4.1 4.2 4.3 4.4 4.5 4.6

In Fork Length (mm)

Figure 12. Initial and ﬁnal (after two weeks) In length-In weight relationships for
the two treatment groups in Growth Experiment 11 (Level 0 = zero ration
treatment; Level 1 was fed).

85
Discussion

My results are similar to the results of Weatherley and Gill (1981), in a study of
the growth and energetic status of juvenile rainbow trout held at 12°C and exposed to
differing periods of starvation (0, 3, and 13 weeks). In that study, juvenile trout had an
initial mean length of 95 mm and an initial mean weight of 10.74 g. A group fasted for
three weeks had a mean weight loss of 14.5%, while the 13-weck zero ration group had
an average weight loss of 32.5% by the end of the treatment. Recovery growth (at ad
libitum rations) of the starved groups was monitored for 15 weeks, and previous history
of starvation did not negatively affect subsequent growth rates, which equaled or
exceeded that of control (fed) groups. Experiment 11 in my study, though much less
intensive than the above, produced similar results: even when ﬁsh were exposed to severe
conditions of food limitation, they recovered quickly when food availability improved.

Juvenile chinook sahnon fasted for two weeks had ﬁnal length-weight regression
slopes signiﬁcantly higher than 3.0 (Figures 11 and 12). In contrast, experimental groups
that were fed had regression slopes closer to 3.0. The steep slope in fasted ﬁsh was
contrary to expectations that the slope would remain similar or decrease but the intercept
would shift down due to the loss of weight at all sizes. The result can be explained by the
fact that smaller ﬁsh within fasted groups were affected more greatly by fasting than
larger ﬁsh due to higher metabolic requirements, causing a disproportionately greater loss
of weight and energy in smaller salmon. The more severe effect of starvation on smaller

ﬁsh has been demonstrated in overwintering Colorado squawﬁsh (Thompson et al. 1991).

APPENDIX E

APPENDIX E

Experiments on Postmortem Changes in Length and Weight

Purpose/Methods

During April and May 1996, time constraints made it necessary to put sacriﬁced
salmon in bags of water kept on~ ice, without ﬁrst measuring live length and weight for
each individual. Fish designated for body water analyses were stored in a cooler for
varying periods while I collected samples at other sites, and were weighed and measured
at the end of the day prior to being placed into frozen storage. Preliminary observations
showed that this method of interim preservation caused signiﬁcant changes in the length-
weight relationship and estimates of condition, thus making it difﬁcult to compare data
from these samples with data on live measurements. I conducted two experimental trials
to track the changes in length and weight of sacriﬁced hatchery salmon kept in coolers up
to twelve hours after death, and used the results to develop correction factors for delayed
ﬁsh measurements.

The two separate trials for this experiment used Coleman National Fish Hatchery
fall-run chinook (the same ﬁsh from Growth Experiment 11 outlined in Appendix D).
Within atrial, three treatment groups with two levels each (two separate tanks, with
different feeding histories) were used. In the ﬁrst group, ﬁsh were sacriﬁced, then
immediately measured and weighed. Individuals were placed in bags of water which
were kept in an ice-ﬁlled cooler. The ﬁsh were weighed and measured every two to four
hours, for up to twelve hours after death. The second group was sacriﬁced, measured and
weighed, then placed in bags of water and put into a freezer. Fish from a third group

were placed in bags of water left in ambient air temperature conditions and were

86

87

measured every two to four hours until decomposition made it difficult to further handle
the bodies for measurement. At the end of sampling, ﬁsh from cooler and ambient
treatments were also frozen. All ﬁsh were thawed at a later date, their thawed
measurements were taken, and the ﬁsh were dried to a constant weight at 70°C.

Throughout the experiment, ﬁsh were identiﬁed and tracked individually.

Results

After death, ﬁsh held in bags of water in coolers tended to shrink in length and
increase in weight, resulting in an inﬂation of the condition factor K. The changes were
not constant over the duration that postmortem length and weight were tracked, however.
During the ﬁrst three hours after death, changes in length and weight were small, with
less than a 1% change from the original measurements. From 4 to 11 hours postmortem,
lengths decreased by 0.56 to 2.56% and weights increased by 1.41 to 6.03%. Comparing
live measurements to corresponding thawed measurements, ﬁsh decreased in length by
4.58% on average after being frozen and thawed, and increased in weight by 6.68% after
ﬁ'eezing and thawing.

Equations used to convert delayed postmortem measurements of length and
weight taken on ﬁeld-caught ﬁsh in April and May to estimates of live length and weight
are given by the following equations:

Live length = delayed length/(1 — 0.0156)
Live weight = delayed weight/(l + 0.0287)
The correction equations are based on the weighted averages of length or weight changes

recorded from 4 to 1 1 hours postmortem, a 1.56% decrease in length and a 2.87%

88

increase in weight. These percentages correspond to an assumed average delay in

measurement of about 7 hours postmortem.

Discussion

Published studies on the effect of postmortem changes due to this method of
interim preservation are absent from the literature. However, Sayers (1987) studied the
effects of freezing on the carcasses of adult and juvenile bloaters Coregonus hoyi. As in
my study, ﬁsh frozen in water and later thawed signiﬁcantly decreased in length, and
signiﬁcantly increased in weight. To correct for changes in length, Sayers (1987)
suggested simply adding 5 mm to the length of ﬁozen bloaters. Application of a
correction factor to the lengths and weights of salmon captured for my study in April and

May helped to improve interpretation of results.

LIST OF REFERENCES

LIST OF REFERENCES

Adams, SM. 1990. Status and use of biological indicators for evaluating the effects of
stress on ﬁsh. American Fisheries Society Symposium 8: 1-8.

Allan, J .D. and E. Russek. 1985. The quantiﬁcation of stream drift. Canadian Journal
of Fisheries and Aquatic Science 42: 210-215.

Arnett, RH, Jr. 1985. American Insects: A Handbook of theInsects of America North of
Mexico. New York: Van Nostrand Reinhold.

Beaucharnp, D.A., D.J. Stewart, and G.L. Thomas. 1989. Corroboration of a
bioenergetics model for sockeye salmon. Transactions of the American Fisheries
Society 118: 597-607.

Becker, CD. 1973. Food and growth parameters of juvenile chinook sahnon,
Oncorhynchus tshawytscha, in central Columbia River. Fishery Bulletin 71(2):
387—400.

Beeman, J.W., D.W. Rondorf, and ME. Tilson. 1994. Assessing smoltiﬁcation of
juvenile spring chinook salmon (Oncorhynchus tshauytscha) using changes in

body morphology. Canadian Journal of Fisheries and Aquatic Science 51: 836-
844.

Bolger, T. and PL. Connolly. 1989. The selection of suitable indices for the
measurement and analysis of ﬁsh condition. Journal of Fish Biology 34: 171-182.

Bowles, DE. and RA. Short. 1988. Size composition of invertebrate drift and ﬁsh
predation in a Texas stream. The Southwestern Naturalist 33(2): 177-184.

Brittain, J .E. and T.J. Eikeland. 1988. Invertebrate drift—a review. Hydrobiologia 166:
77-93.

Brown, R.L. 1991. Bioengineering problems in river systems of the Central Valley,
California. American Fisheries Society Symposium 10: 19-31.

89

90

Brodeur, RD. 1992. Factors related to variability in feeding intensity of juvenile coho
salmon and chinook salmon. Transactions of the American Fisheries Society 121:
104-114.

Brodeur, R.D., R.C. Francis, and W.G. Pearcy. 1992. Food consumption of juvenile
coho (Oncorhynchus kisutch) and chinook salmon (0. tshauytscha) on the
continental shelf off Washington and Oregon. Canadian Journal of Fisheries and
Aquatic Science 49: 1670-1685.

Buer, K. 1984. Middle Sacramento River spawning gravel study. Red Bluff, California:
California Department of Water Resources Northern District.

Cada, G.F., J.M. Loar, and M.J. Sale. 1987. Evidence of food limitation of rainbow and
brown trout in southern Appalachian soft-water streams. Transactions of the
American Fisheries Society 116: 692-702.

California Department of Water Resources. 1994a. Use of alternative gravel sources for
ﬁshery restoration and riparian habitat enhancement in Shasta and Tehama
Counties, California. Red Bluff: DWR Northern District.

California Department of Water Resources. 1994b. Sacramento River bank erosion
investigation progress report. Red Bluff: DWR Northern District.

Carl, L.M. 1984. Chinook salmon (Oncorhynchus tshawytscha) density, growth,
mortality, and movement in two Lake Michigan tributaries. Canadian Journal of
Zoology 62: 65-71.

Caulton, M.S. and E. Bursell. 1977. The relationship between changes in condition and
body composition in young T ilapia rendaIIi Boulenger. Journal of Fish Biology
11: 143-150.

Clark, RD. and J .J . Strong. 1991. Factors affecting performance of the Glenn-Colusa
ﬁsh screen. American Fisheries Society Symposium 10: 249-255.

Cone, RS. 1989. The need to reconsider the use of condition indices in ﬁshery science.
Transactions of the American Fisheries Society 118: 510-514.

F ield-Dodgson, M.S. 1988. Size characteristics and diet of emergent chinook salmon in
a small, stable New Zealand stream Journal of Fish Biology 32: 2740.

Fisher, F.W. 1992. Chinook sahnon Oncorhynchus tsauytscha growth and occurrence in
the Sacramento-San Joaquin river system. Red Bluff, CA: California Department
of Fish and Game, Inland Fisheries Division.

91

Fisher, F.W. 1994. Past and present status of Central Valley chinook sahnon
Conservation Biology 8(3): 870-873.

Garcia, A. 1989. The impacts of squawﬁsh predation on juvenile chinook sahnon at Red
Bluff Diversion Dam and other locations in the Sacramento River. Red Bluff
CA: US. Fish and Wildlife Service Fisheries Assistance Office. USFWS Report
No. AFF/FAO-89-05.

Gardiner, W.R. and P. Geddes. 1980. The inﬂuence of body composition on the survival
of juvenile salmon. Hydrobiologia 69: 67-72.

Godinho, AL. 1997. Weight-length relationship and condition of the characiform
Triportheus guentheri. Environmental Biology of Fishes 50: 319-330.

Goede, RW. and BA Barton. 1990. Organismic indices and an autopsy-based
assessment as indicators of health and condition of ﬁsh. American Fisheries
Society symposium 8: 93-108.

Greene, S. 1992. Estimated winter-run chinook salmon salvage at the State Water
Project and Central Valley Project Delta pumping facilities. Memorandum to R.
Brown, 5/8/92. California Department of Water Resources.

Hard, U. 1986. Production and yield of juvenile chinook salmon in two Alaskan lakes.
Transactions of the American Fisheries Society 115: 305-313.

Hartman, KJ. and SB. Brandt. 1995. Estimating energy density of ﬁsh. Transactions
of the American Fisheries Society 124: 347-355.

Hayes, DB. and W.W. Taylor. 1994. Changes in the composition of somatic and
gonadal tissues of yellow perch following white sucker removal. Transactions of
the American Fisheries Society 123: 204-216.

Healey, MC. 1983. Coastwide distribution and ocean migration patterns of stream- and
ocean-type chinook sahnon, Oncorhynchus tshawytscha. Canadian F ield-
Naturalist 97(4): 427-433.

Herbinger, C.M., and G.W. Friars. 1991. Correlation between condition factor and total
lipid content in Atlantic salmon, Salmo salar L., parr. Aquaculture and Fisheries
Management 22: 527-529.

Higgs, D.A, J.S. Macdonald, C.D. Levings, and BS. Dosanjh. 1995. Nutrition and
feeding habits in relation to life history stage. In C. Groot, L. Margolis, and WC.
Clarke (eds.), Physiological Ecology ofPaciﬁc Salmon. Vancouver, British
Columbia: UBC Press.

92

Hoar, W.S. 1976. Smolt transformation: evolution, behavior, and physiology. Journal
of the Fisheries Research Board of Canada 33: 1234-1252.

Hyslop, E.J. 1980. Stomach contents analysis—a review of methods and their
application. Journal of Fish Biology 17: 411-429.

Jonas, J.L., C.E. Kraft, and TL. Margenau. 1996. Assessment of seasonal changes in
energy density and condition in age-0 and age-1 muskellunge. Transactions of
the American Fisheries Society 125(2): 203-210.

Keeley, ER. and J .W.A. Grant. 1997. Allometry of diet selectivity in juvenile Atlantic
salmon (Salmo salar). Canadian Journal of Fisheries and Aquatic Science 54:
1894-1902.

Kjelson, M.A. 1993. Emigration. In Notes and Selected Abstracts from the Workshop
on Central Valley Chinook Salmon, University of California at Davis, 4-5
January 1993. Davis: University of California Department of Wildlife and
Fishery Biology. 46 pp.

Kjelson, M.A. and PL. Brandes. 1989. The use of smolt survival indices to quantify the
effects of habitat changes on salrnonid stocks in the Sacramento-San Joaquin
rivers, California. P. 100-115. In C.D. Levings , L.B. Holtby, and M.A.
Henderson (eds), Proceedings of the National Workshop on Eﬂects of Habitat
Alteration on Salmonid Stocks. Canadian Special Publication in Fisheries and
Aquatic Science 105.

Kjelson, M.A., P.F. Raquel, and F .W. Fisher. 1982. Life history of fall-run juvenile
chinook salmon, Oncorhynchus tshawytscha, in the Sacramento-San Joaquin
estuary, California. p. 393-411. In V.S. Kennedy (ed), Estuarine Comparisons.
New York: Academic Press.

Kolok, AS. and D.W. Rondorf. 1987. Effect of differential gastric evacuation and
multispecies prey items on estimates of daily energy intake in juvenile chinook
sahnon. Environmental Biology of Fishes 19(2): 131-137.

LeCren, ED. 1951. The length-weight relationship and seasonal cycle in gonad weight
and condition in the perch (Percaﬂuviatilis). Journal of A nimal Ecology 20:
201-219.

Littell, R.C., R.J. Freund, and RC. Spector. 1991. SAS System for Linear Models. 3rd
edition. Cary, NC: SAS Institute.

Love, RM. 1970. The Chemical Biology of Fishes. New York: Academic Press.

93

Merritt, RW. and KW. Cummins (eds). 1984. An Introduction to the Aquatic Insects of
North America. 2MI edition. Dubuque, Iowa. KendalVHunt Publishing Co.

Moore, T. L. 1997. Condition and feeding of juvenile chinook salmon in selected
intermittent tributaries of the upper Sacramento River. M. S. Thesis, California
State University, Chico, CA. 66 pp.

Moyle, PB. 1976. Inland Fishes of California. Berkeley: University of California
Press.

National Marine Fisheries Service. 1994. Endangered and threatened species; status of
Sacramento River winter-run chinook sahnon. Federal Register 440, Volume
59(2): January 4, 1994.

Pennak, R.W. 1978. Freshwater Invertebrates of the United States. 2"d edition. New
York: J.Wiley and Sons. 803 pp.

Ricker, W.E. 1979. Growth rates and models. In Fish Physiology Volume 8. New
York: Academic Press.

Rondorf, D.W., G.A. Gray, and RB. Fairley. 1990. Feeding ecology of subyearling
chinook sahnon in riverine and reservoir habitats of the Columbia River.
Transactions of the American Fisheries Society 119: 16-24.

Sagar, PM. and G. J. Glova. 1987. Prey preferences of a riverine population of juvenile
chinook salmon, Oncorhynchus tshawytscha. Journal of Fish Biology 31: 661-
673.

Sagar, PM. and G]. Glova. 1988. Diel feeding periodicity, daily ration, and prey
selection of a riverine population of juvenile chinook salmon, Oncorhynchus
tshawytscha (Walbaum). Journal of Fish Biology 33: 643-653.

Saﬁan, P. 1992. Theoretical analysis of the weight-length relationship in ﬁsh juveniles.
Marine Biology 112(4): 545-551.

Salam, A. and P.M.C. Davies. 1994. Body composition of northern pike (Esox lucius L.)
in relation to body size and condition factor. Fisheries Research 19: 193-204.

Sayers, R.E., Jr. 1987. Effects of freezing in and out of water on length and weight of
Lake Michigan bloaters. North American Journal of Fisheries Management 7:
299-301.

SAS Institute, Inc. 1989. SAS/STAT User ’s Guide, Version 6. 4‘h Edition, Volume 2.
Cary, NC: SAS Institute, Inc. 846 pp.

94

Schaﬁter, R.G., P.A. Jones, and J.G. Karlton. 1982. Sacramento River and tributaries
bank protection and erosion control investigation: evaluation of impacts on
ﬁsheries. Final report. Red Bluff, CA: California Department of Fish and Game.

Schlosser, U. 1991. Stream ﬁsh ecology: a landscape perspective. Bioscience 41(10):
704-712. -

Springer, T.A., B.R Murphy, S. Gutreuter, RO. Anderson, L.E. Miranda, D.C. Jackson,
and RS. Cone. 1990. Properties of relative weight and other condition indices.
Transactions of the American Fisheries Society I 19:1048-1058.

Strauss, RE. 1979. Reliability estimates for Ivlev’s Electivity Index, the Forage Ratio,

and a proposed linear index of food selection. Transactions of the American
Fisheries Society 108: 344-352.

Thompson, J.M., E.P. Bergersen, C.A. Carlson, and LR. Kaeding. 1991. Role of size,
condition, and lipid content in the overwinter survival of age-0 Colorado
squawﬁsh. Transactions of the American Fisheries Society 120: 346-353.

Usinger, R.L. 1956. Aquatic Insects of California. Berkeley: University of California
Press.

Vogel, DA. and KR. Marine. 1991. US. Bureau of Reclamation, Central Valley
Project, Guide to Upper Sacramento River Chinook Salmon Life History.
Redding, CA: CH2M Hill.

Waters, T.F. 1972. The drift of stream insects. Annual Review of Entomology 17: 253-
272.

Wainwright, PC. and BA. Richard. 1995. Predicting patterns of prey use from
morphology of ﬁshes. Environmental Biology of Fishes 44: 97-113.

Weatherley, AH. and HS. Gill. 1981. Recovery grth following periods of restricted
rations and starvation in rainbow trout Salmo gairdneri Richardson. Journal of
Fish Biology 18: 195-208.

Weatherley, AH. and HS. Gill. 1983. Protein, lipid, water and caloric contents of
immature rainbow trout, Salmo gairdneri Richardson, growing at different rates.
Journal of Fish Biology 23: 653-673.

Wootton, R]. 1992. Ecology of T eleost Fishes. London: Chapman and Hall.

"111111111111111i