United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97330
EPA-600 3-79-043
April 1979
Research and Development
Sediment  Particle
Sizes Used by
Salmon for
Spawning with
Methods for
Evaluation

-------
                 RESEARCH REPORTING SERIES

 Research reports of the Office of Research and Development, U.S. Environmental
 Protection Agency, have been grouped into nine series. These nine broad cate-
 gories were established to facilitate further development and application of en-
 vironmental technology. Elimination of traditional  grouping was consciously
 planned to foster technology transfer and a maximum interface in related fields.
 The nine series are:

      1.  Environmental Health Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental  Studies
      6.  Scientific  and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research, and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

 This report has been assigned to the ECOLOGICAL RESEARCH series. This series
 describes research on the effects of pollution on humans, plant and animal spe-
 cies, and materials.  Problems are assessed for their long- and short-term influ-
 ences. Investigations include formation, transport, and pathway studies to deter-
 mine the fate of pollutants and their effects. This work provides the technical basis
 for setting standards to minimize undesirable changes in living organisms in the
 aquatic, terrestrial, and  atmospheric environments.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                             EPA-600/3-79-043
                                             April  1979
   SEDIMENT PARTICLE SIZES  USED BY  SALMON  FOR
      SPAWNING WITH METHODS FOR EVALUATION
                       by
                William S.  Platts
Intermountain Forest and Range Experiment Station
                 Forest Service
         U.S.  Department of Agriculture
               Ogden, Utah  84401

                       and

     Mostafa A. Shirazi and Donald H. Lewis
           Freshwater Systems Division
   Corvallis Environmental Research Laboratory
      U.S. Environmental Protection Agency
            Corvallis, Oregon  97330
   CORVALLIS  ENVIRONMENTAL  RESEARCH  LABORATORY
       OFFICE OF  RESEARCH AND  DEVELOPMENT
      U.S.  ENVIRONMENTAL PROTECTION  AGENCY
            CORVALLIS,  OREGON   97330

-------
                                  DISCLAIMER

     This  report has  been  reviewed by  the Corvallis  Environmental  Research
Laboratory, U.S.  Environmental  Protection  Agency,  and approved  for publica-
tion.   Mention  of  trade names  or  commercial  products  does  not  constitute
endorsement or recommendation for use.
                                     ii

-------
                                   FOREWORD

     Effective regulatory and enforcement actions by the Environmental Protec-
tion Agency  would be  virtually impossible without  sound scientific  data  on
pollutants and their  impact on environmental  stability and human health.  Re-
sponsibility for  building  this data base has been assigned to EPA's Office of
Research and Development and its 15 maior field installations, one of which is
the Corvallis Environmental  Research Laboratory (CERL).

     The primary  mission of  the  Corvallis Laboratory is research  on the ef-
fects  of  environmental  pollutants  on  terrestrial,  freshwater,   and  marine
ecosystems; the behavior,  effects  and control of pollutants in lake and river
systems; and  the  development  of  predictive models on the  movement of pollu-
tants in the biosphere.

     This  report  addresses  a non-point source  pollution problem  of special
regional interest to  the  Pacific  Northwest.   Salmon in  this  region spawn in
head waters  subjected to  intensive silvicultural  activities.   Sediment laden
runoffs from disrupted land surfaces could degrade the spawning habitat in the
adjacent stream.   The  identification  and analysis of such  habitats is a sub-
ject of this report.

                                       James C.  McCarty
                                       Acting Director,  CERL

-------
                                   ABSTRACT

     The size composition of substrates used by Chinook salmon for spawning in
the  South Fork  Salmon  River,  the  main Salmon River  and tributaries  of  the
Middle Fork  Salmon  River,  Idaho,  was determined.   Substrates used by resident
trout  were  analyzed  for streams  in the BoiSe and Payette  River  drainages.
These  analyses  were made over  time to determine particle  sizes  preferred by
spawning  salmon,  yearly differences  in sizes used by these  salmon,  the size
differences  used  by spring  and  summer Chinook salmon,  and differences between
channel  sediments used by  Chinook salmon  for  spawning  and  those  substrates
occupied by trout.

     The use of  the geometric  mean particle diameter method is presented as a
companion  measurement  to  "percent  fines"  for a  more  complete  analysis  of
sediments  used  for  spawning.  The geometric mean  particle diameter  is  more
adaptive to statistical analysis than the more common method of using "percent
fines."  The geometric  mean diameter of the  sediment particle  size distribu-
tion  is  used for analyzing channel sediments.  The relationship between  the
geometric mean particle  diameter  and "percent fines,"  substrate permeability,
and substrate porosity  is  established.   The strongest correlation between the
two methods  of  analysis,  "percent fines" and geometric mean diameter, was  for
fine sediments below 0.88 in (2 mm) in particle size.

     Chinook salmon selected sediments  for spawning that were  mainly between
 28 and  .79  in  (7.0 to 20 mm)  in geometric mean particle diameter,  regardless
of stream  selected.  This  is a  narrow range considering that the mean particle
diameters  for  streambed sediments  available  for  chinook  salmon  to  spawn  in
vary from less than 0.02 in (.5 mm) to well over 3.94 in  (100 mm).  The compo--
sition of spawning sediments selected by chinook salmon each year between 1966
and 1976  were  quite uniform.  Sediments  used for spawning in  the  South Fork
Salmon River decreased  in  particle size in a downstream direction.   Geometric
mean diameters 35 miles  below  the headwaters averaged .35 in (8.8 mm); parti-
cles 10 miles below the headwaters averaged  .58 in (14.7 mm).
                                      iv

-------
                                   CONTENTS
                                                                        Page
Abstract	iv
Introduction 	   1
Study Area	3
Methods for Describing Spawning Sediments  	   8
Procedures	16
Results and Conclusions  	  18
References	30

-------
                                 INTRODUCTION

     Most stream fishes require channel sediments having a variety of particle
size mixes  for survival.   This is especially true for salmonids which deposit
their  eggs  in  sediments  of a  particular  size class.  However,  studies have
demonstrated that the redd sediments must be of the proper particle size class
and  composition for high  embryo survival.  Large  increases  in fine sediment
loads  into  stream  channels can  create  intolerable channel  modifications  in
salmonid spawning areas  (Platts and Megahan 1975).  Hall  and Lantz (1969),  in
their  Alsea,  Oregon logging  studies,  found that an increase of  5  percent  in
fine sediment  smaller than 0.033  in (.83  mm) in  diameter  in redds decreased
survival of emergent coho  salmon fry (Oncprhynchus  kisutch Walbaum).  Other
authors  have  demonstrated that fine  sediment  particlesdeposited  in  the
streambed  reduce permeability  and thus  cause  higher  egg-to-fry mortality
(McNeil  and Ahnell   1964).  The literature  supports  the  statement  that fine
sediments can  limit fish productivity.   However,  there is a dearth of litera-
ture identifying and evaluating the effect of different  mixtures of sediment
sizes on fish health and survival in the actual stream environment.

     During their evolutionary period  salmon and trout adapted to the natural
channel sediments.  Salmonids need sediment for spawning,  rearing their young,
and providing  for their food.  However, the mix of sediment particle sizes for
optimum  fish   productivity  is  not  clear.    Probably  no  single particle size
group (i.e., boulder, rubble, gravel or fine sediment) will  create the type of
environment salmonids require  for growth and survival.  More likely, a complex
mixture  of  sediment  sizes is needed in  combination with  certain hydraulic
conditions to provide the ideal channel  environment.

     Since  streams  offer a wide variety  of sediment sizes,  salmon entering
virtually any  river area  can select any  particle size for spawning.  Stream
channel substrates  are  available from 100  percent fine sediments to channels
that are  all   boulder  or rubble.  The  fish  seldom find channels  composed en-
tirely  of   gravel  because gravels  are  usually mixed  with  fine  sediment and
small  rubble.   However,  some  hydraulic environments such as  heads  of riffles
may  sort  out  most  of the fine  sediments.   Throughout their  evolution, it  is
probable that  those salmon that spawned in fine sediments,  rubble or boulders
failed  to   survive  as well as salmon that spawned in predominantly gravel.
Somewhere between  the  extremes of  fine sediment  and  rubble  is the optimum
composition  composed mainly  of  gravel  mixed with smaller  amounts  of fine
sediment and small rubble.

     Most  salmon become  riffle spawners  because  embryo  survival  requires
specific conditions such as  water velocities, water  depths, sufficient dis-
solved  oxygen  and  embryo  metabolic waste  removal.   The  hydraulic conditions
that build  and maintain these spawning riffles are widespread and persistent
enough so that through time and over space, salmon were able to develop  habit-

-------
 ual spawning areas.  Although there  may  be some minor changes in riffle loca-
 tion from year to  year they are usually  slight enough to cause no problems to
 salmon homing.   Thus, each year  salmon usually seek a predetermined area  for
 deposition of  their eggs.   Salmon  usually select  areas  where  the  hydraulic
 controls on the stream  channel  provide  a substrate almost devoid of boulders
 because fish can't move them,  low in fine  sediments  because of  the need  for
 subsurface water permeability,  and high  in gravel  and small  rubble which they
 can form into a cover  that protects  the  eggs and alevins.  This particle size
 distribution  provides an egg cover that  will  withstand most  of the velocities
 the  stream  exerts  without  sediment  movement  damaging  the  embryos.   It  is
 interesting  that the fish  do not choose  channel substrates  completely  devoid
 of fine sediments,  even though  such  areas  exist.   Thus,  it  is  possible that
 fine sediments in  the  correct  amounts  can be  important to  embryo  survival.
 Possibly,  and this  is based only on intuitive  thinking, proper amounts of fine
 sediments  could protect the eggs  from predators,  keep  organic materials  in  the
 stream flow  from settling  on the eggs,  keep eggs  from being buffeted by high
 sub-surface  flows,  and help keep eggs  and  alevins  in  the  substrate  during
 floods  until time for their emergence.

      A  confounding  factor  to us  in  determining why  salmon  choose a certain
 spawning area  is that the  quality of  the surrounding  rearing environment that
 guarantees  survival  of  their young must  also be  a major factor  in spawning
 site  selection.  We believe salmon select  spawning sites by ocular  selection
 of desirable sediment  size  classes,  a  feel  for the  required surface  water
 velocities to  drive the needed subsurface  flows  for the  embryos  and alevins,
 and a strong  homing  instinct that places them in an area in  which their young
 have a good chance to survive.

     Although  salmonids have survived sedimentation  from  the watershed over
 the past million years,  the literature indicates  that their ability to cope
 with  sudden  increases  in  channel sedimentation  may not be  very  good.  Thus
 certain  questions  relating to watershed management  need better answers:  Have
 stream  channel sediment  size  classes changed because of man's influences?   Has
 there  been a  resulting change   in  the   spawning  success  of salmonids?   Can
 salmonids  adjust  to  changes  in   the  quality  of channel  sediments over time?
 Have  fish  evolved  to survive only within narrow ranges of channel sedimenta-
 tion  or can  they   survive  under  wide variations?  Do  we know what channel
 sediment particle   sizes  and particle size  composition  fish need  for good
 health  and survival?  If so, how closely do we need to be able to  measure this
 composition for optimum  fisheries  management?

     This  report contributes some answers  for  these  questions by describing
channel  sediment   particle  size  mixtures  Chinook  salmon   (Oncorhynchus
tshawytscha Walbaum)  use for spawning over broad streambed areas.  Methods  for
the analysis and evaluation of those sediments selected  are discussed.

-------
                                  STUDY AREA

     The  Salmon  River drainage supports most of the Chinook salmon that enter
Idaho  to  spawn.   These  waters are usually low in  mineral  content because of
the  predominance  of granitic bedrock.  A major part of the Salmon River water-
shed is within the 16,000-square-mile (6,150 km2)  Idaho Batholith, an area of
granitic  bedrock much of which is characterized by  steep slopes, erosion-prone
soils, and  severe climatic stresses.  Soil disturbances, such as those associ-
ated with  logging  and  road  construction, can  accelerate soil  erosion  many
times  over  natural rates  on  such  lands.   Part  of the  Salmon River drainage
lies in the Belt Series which is not granitic, and  other bedrock types such as
volcanics and sedimentaries occupy relatively small sections.

     The  Salmon  River drainage (Figures 1 and 2) ranges from over 12,000 feet
(3600  m)  above sea level in headwater areas to about 1500 feet (450 m) at its
confluence  with  the  Snake River.   Most of the  spawning  areas  occur between
5000 and 7000  feet (1650-2100 m), which corresponds to some important sediment
dumps  formed  by  glaciers during the  Pleistocene  epoch.   These streams formed
themselves  in  these extensive  Pleistocene glacial deposits.  This sediment was
transplanted from  higher elevations by glaciers and deposited in moraines and
outwash trains.  Subsequently, stream channels have reworked this sediment and
evolved to their present morphology in quasi-equilibrium with climatic change.
Part  of  the   reason  Chinook  salmon  and  steel head  trout  (Salmo  gairdneri
Richardson) spawn  and rear on these glacial  dumps  is  because of the abundant
supply of suitable  sediment  particle sizes at elevations  creating cool  water
temperatures.

     The Boise River drainage (Figure 3) ranges from over 10,000 feet (2048 m)
to about  2600  feet (792 m) at  its confluence with the Snake River.  This river
also drains  an  area of granitic bedrock.

     Thi& study  was mainly conducted in the Salmon  River drainage including
its  two  major tributaries, the  South  Fork  Salmon River  and  the Middle  Fork
Salmon River.  The  South Fork drains a  1,270-square-mile  (660 km2) watershed
representative of the  forested  mountainous  terrain  found in  central  Idaho.
The Middle  Fork  is  a larger drainage that  depends  on its tributaries for the
spawning of Chinook  salmon and steel head trout.   The South  Fork channel  con-
tains  the  necessary sediment  particle  sizes required for  spawning while the
Middle Fork channel does not.   The stream power in the Middle Fork is too high
to allow  sufficient quantities  of  gravel  and fine sediment to  remain in the
channel.   Therefore,  salmon  move   into  the  tributaries to  find the  size of
channel materials they need for spawning.  In the South Fork there are channel
reaches with low  enough stream power to allow accumulation and containment of
gravels and fine  sediment.  However,  salmon use the tributaries  in the South
Fork much  less than  in the  Middle Fork.   The main  river has  large channel
areas composed of  gravel and fine sediments.

-------
                                                • Obsidian
                                                    Salmon River
                                                     Alturas Lake ,
                                                      Creek
                                                    Perkins Lake
                                                 Alturas Lake
                   1    2
                    Miles
Figure 1.  Study sites in the headwaters  area of  the Salmon River.

-------
                                       almon River
Secesh
River
                     . Fork South Fork
                                  Salmon River
                                                  Middle Fork
                                                 Salmon River
                          Johnson Creek

                                     Ik Creek
           Stolle
          Meadows

             SouthlFork
            Salmor/River
                                                      1  3
                                                     Miles
 Figure 2.   Streams  studied  in the Salmon  River  drainage.

                                   5

-------
                                                          SouthsFork
                                                          Boise* River
                                                         Anderson Ranch
                                                          Reservoir
                                  Miles
Figure  3.   Study  areas in  the Squaw Creek and South Fork Boise  River drainages.

                                          6

-------
     Only summer chinook  use  the South Fork Salmon River for spawning;  spring
Chinook are  the primary  species  using  the  Middle Fork drainage and  the  main
Salmon River.

-------
                   METHODS FOR DESCRIBING SPAWNING SEDIMENTS

     Sediments with different particle size compositions can be compared using
the  respective  particle size-cumulative distribution  curves.   Data  for these
curves can be obtained through standard particle size analysis with percent of
sediment  by  weight that is  finer  or coarser than a given  sieve  size plotted
against  that opening  size.  The  use  of  logarithmic  abscissa  is  desirable
because natural  sediments  have  an extremely wide range of grain sizes spread-
ing  over  three   or  more cycles  (i.e.,  factors of  10).   Furthermore, natural
sediments frequently exhibit lognormal distributions, i.e.,  when the logarithm
of the particle  size (instead of the particle  size  itself) is used, the dis-
tribution  is nearly  normal.  Such  nearly  lognormal distributions  show more
symmetrical patterns on semi-log papers and their cumulative distributions are
close to straight lines on log probability papers.

     Following a  conventional statistical  approach,  it is possible to compare
two  different  sediment samples  by some representation of  the particle size-
cumulative distribution curves in place of the entire curves.  For example, if
the  curves  were  truly  lognormal,  the means  and variances  of the cumulative
distributions could  be the  only information needed to define the curve.  If
the  curves are  skewed,  additional  information is required  to  show, the skewed
effects.   The use of mean  and variance simplify  the comparison considerably,
even when the distributions are not truly lognormal.

     Numerical  integration procedures for calculation of the mean, variance or
skewness are available  and  can  be applied to  the data once the particle size
cumulative  distribution is  known.   Because  this is  tedious, graphical  ap-
proaches are better suited for estimating such standard parameters as the mean
and  variance.  For example,  the  median  particle size, d50,  is picked up from
the  graph of  the cumulative  distribution  curve directly  to represent  the
particle diameter for  which 50  percent dry weight of  the sediment is coarser
or  finer.   If the  distribution  is  lognormal,  this is  exactly equal  to the
geometric mean  of the  distribution.   For  normal distributions,  one standard
deviation on  either side of the mean diameter is approximately  d16  and d84,
respectively, and the  2.5  and 97.5 percentiles are two standard deviations on
either side of the mean.  Once the cumulative distribution curve of a sediment
composition is plotted, all  such parameters can be directly picked up from the
curves with no further calculation.

     When  the  distribution  is   not  symmetrical  or  lognormal, Innman (1952)
following the classic  work  of Otto, recommended using the  geometric mean of
the  particle diameters  corresponding to the 16th and  84th  percentiles (i.e.,
d16  and d84).  This  has  now become a standard  procedure  (Vanoni  1977).  That
is,  the geometric mean  diameter  obtained from d16 and d84 is used even if the
distribution deviates  from  lognormal.   The geometric  mean diameter,  d ,  is
obtained from d16 and dS4 as follows:

                                      8

-------
An estimate of the standard deviation, a , is obtained by:
                                       y
     The  mean diameter  d  is a  useful  measure,  since  it can be manipulated
algebraically  (Innman 195%).   For example, the  mean  particle size of several
combined  samples  is equal to the  average  of  the means of those samples; this
is not true for the  median, i.e. , d50.

     It  is  interesting to note that small values of o  are usually associated
with  small   d   values,  frequently  in  large  streams. ^In the case  of coarse
sediments,  i.e., when d   is relatively large, the geometric standard deviation
is  also  relatively larTje (Bogardi  1974),   This suggests  that  d  is  both a
convenient and sufficient way  to describe substrate composition.    ^

     Innman  (1952)  reported on Yule and  Kendall's  calculations  showing that,
for a  normal curve, sampling error is greatly  increased below 5 and above 95
percentiles  of the distribution.   The  errors  are tolerable  for 16  and 84
percentiles.   This  is an important practical consideration  when sampling for
substrate composition and will be discussed further.

     Fishery  scientists  have characterized stream  channel  sediments  by "per-
cent fines", which  is defined as  the mass  fraction below a suitable selected
particle  size.   There has  been  considerable debate  (Iwamoto  et al .  1978) on
the  choice   of the  suitable  particle size  as   it  relates  to egg  and alevin
mortality.   Common  particle sizes  fishery scientists use to identify "percent
fines" are  .03 in (0,83 mm), .13 in (3.3 mm), .19 in (4.7 mm), and .25 in (6.3
mm).  This has rendered comparison of research results difficult if not impos-
sible.   There  are reasons why fishery scientists have used different particle
size  limits to  define  fine  sediments.   One reason  is  that  salmon  spawning
areas in the Pacific Northwest exhibit different particle size graduations and
researchers  concerned with  these  areas  observe different  dominant  features
affecting embryo survival.  A second reason is that mortality has been intrin-
sically  associated  with  excess fine  particles  because of  (a)  the  adverse
effects very fine  particles  have  on permeability  and (b)  the  entrapment of
embryos that can  be caused by presence of particles of intermediate fineness,
say, 2  mm- 6  mm.

     Note the  difference  in  emphasis  between  the "percent  fines"  and  the
percentile  approach.   With  the "percent  fines"  the particle  diameter  is se-
lected and then the fraction of the sample which is finer is determined.  With
the percentile method, the percent passing is selected and the sample analyzed
to find  the corresponding  particle diameter.  The  inherent  disadvantages of
the "percent fines" approach is  that  the probability of occurrence  of these
quantities,  e.g., percent by weight less  than .03 in (.83 mm), etc., vary from
one composition  to  another.  This  means  that if the  sediment  composition is
coarse, it  would  be more difficult to evaluate its "percent fine" by sampling
than if  the  sediment composition  is fine.   The percentile  approach  always
results in the  same  sample size.

-------
     For example,  the channel  substrate  in the  Dollar  and Poverty  spawning
areas on the  South Fork Salmon River contains  5 and 14 percent,  respectively,
"percent fines"  less  than  .03  in  (.83 mm).   Therefore,  because of  smaller
"percent fines"  in the Dollar  area, the  sampling  error associated with  its
determination is expected to  be greater than for the Poverty area.   To attain
comparable   accuracies  while  using  the  same sampling procedures  in  the  two
areas,  "percent  fines" in  the Dollar  area must be  based  on  .08  in  (2  mm)
particles where  14% of  the  substrate is finer.  This presents  an  intolerable
paradox; how  do  we know  in advance  what basis (i.e., particle  diameter)  for
percent fines to choose?  Measures based on geometric mean avoid this particu-
lar difficulty because  the  independent  variable,  i.e., the particle  size,  is
not  predetermined  and  it may vary,   as  it actually  would  from one  sediment
sample to another.   Also,  if it is desirable to determine the smaller particle
sizes, thereby placing more  emphasis  on  the amount of fines,  then such quanti-
ties as d16,  ds, d2<5 etc.,  are more suitable  than "percent fines".  These d
levels can  then be  used to determine  degree and causes of mortality in embryos
and alevins.

     The intuitive appeal of  "percent fines" in fisheries studies  stems  from
its long association with impacts on egg survival  (Iwamoto  et  a_K  1978).   As
stated earlier, it  has been  verified  that intergravel flow of water and oxygen
is strongly related to  percent  fines and thus to spawning success.   This  is a
very important and legitimate argument.   Our studies show that as a measure of
intergravel  flow, the  geometric mean  is  at least as  good a measure as "percent
fines".

     Cooper  (1965),  presents data  relating survival  of eyed  sockeye salmon
(Oncorhynchus nerka Walbaum) eggs with intragravel  water flow (Table 1).


      TABLE 1.  RELATION BETWEEN RATE OF WATER  FLOW THROUGH  A GRAVEL BED
                AND THE SURVIVAL OF EYED SOCKEYE EGGS IN THE GRAVEL1.

Apparent velocity (cm/sec)2
through spawning sediments
.0338
.0112
. 00542
.00261
.00136
.000945
.000668
.000389
Percent egg
survival
89.3
78.3
68.3
59.0
36.3
26.5
15.6
1.9

1 Taken from Cooper (1965).
2 Annaront uolnritw onnalc rHcrhavno Hi

\iir\or\ h\/ t-ntal r»»ncc
          sectional  area of voids and solids.

                                      10

-------
     The  positive relationship  is unmistakable  with  these  low  velocities.
However, as velocities  continue  to increase above those reported  here,  there
would be  a level where egg survival  would start to decrease because of  the
pressures  or  buffeting from surface  flows.  Cooper conducted numerous  tests
with gravels of  different  compositions and showed that apparent  velocity is  a
function of gravel porosity and permeability for a given hydraulic head.   That
is,

          V=  f (s, e, p).

     where,

          V =  velocity
          s =  hydraulic head
          e =  porosity
          P =  permeability

     Our analysis of Cooper's  data demonstrates a  strong  correlation between
geometric mean diameter of the appropriate gravels used and  their respective
measured porosity e  and computed permeability p.  These results  are  shown in
Table 2 and Figures 4 and 5.  Accordingly, a single measure of gravel  composi-
tion, d , provides the link between apparent subsurface water velocity and egg
survival.

     Alternatively,  "percent fines" also  are related to porosity with reason-
ably high  correlation,  but a choice has to be made for a proper definition of
"percent fines".  Table 2L shows  the  comparison of porosity  as  a function of
d  ,  percent  fines,  and z£.   The  latter  factor appears in the  definition of
permeability p as reported by Cooper  (1965).  It is the sum of the fraction P
of particles by  weight of  diameter d  divided  by the diameter.  The table was
prepared for  correlation  of  e and p  with d  ,  "percent  fines", etc.  Linear
correlations were best suited  for e but  power functions  of the type Ax  were
more appropriate for p.  This table is not  intended for use of these empirical
correlations.   The degrees  of  fit also are not used here to show conclusively
which are  the best  parameters.  We are dealing  only  with  one set of data and
caution  should  be exercised  in  reading  too  much  into  the result.   Table 3,
however, does serve one important  function, i.e., to show that for this set of
data the geometric  mean particle  diameter  competes in representativeness with
other measures.
                                                        P                P
     There  is  also  an  excellent  correlation  between  Z-j and  p,  since Z^ has
been used  in the definition and calculation of p.  z£ does appear to be a very
good measure, even though it is less conventional ana more difficultpto calcu-
late  than d  .   There  is,   however,  a strong  correlation  between Z-r  and d .
Calculations9are  not presented in the table,  but the coefficient of represen-
tation r2 of a power function of the type Ax  was found to be 0,89.

     The high correlation between  "percent  fines" for the 6.3 mm particle size
and  less  and  p  has  a  curious  explanation associated with the specific nature
of  gravel  used  in  the work.  For the 15  gravel  compositions used,  "percent
fines"  below  6.3 mm averaged  15  percent.   This  is very closely related to d16
used in calculation  of d .  The  rationale  for the strong correlation becomes
                         y
                                       11

-------
           TABLE 2.   POROSITY  AND PERMEABILITY OF  SPAWNING  GRAVEL  AND  ITS  RELATION  TO GRAVEL COMPOSITION1

% Finer than
dg
(cm)


1.61
1.73
3.24
3.35
6.90
1.60
2.09
2.97
4.60
6.26
6.57
4.13
3.12
2.12
1.30
4
(cm"1)


2.66
1.88
3.84
0.77
0.41
4.35
3.93
1.54
0.58
0.40
0.40
1.12
1.93
3.07
4.67

.83
(mm)


2.0
3.5
7.9
1.0
0.4
6.7
5.2
3.5
1.7
0.2
0.2
2.4
4.0
5.2
6.7

6.3
(mm)


19.4
20.9
27.6
9.2
4.4
25.9
18.8
12.6
7.4
4.9
5.0
9.0
12.5
17.5
27.5
Porosity, e

Loose
bed


0.278
0.298
0.233
0.305
0.412
0.235
0.269
0.295
0.316
0.371






Compact
bed


0.200
0.232
0.111
0.254
0.382
0.186
0.235
0.248
0.283
0.334






Mean
e


0.244
0.265
0.172
0.280
0.397
0.211
0.252
0.272
0.300
0.353
0.327
0.278
0.240
0.217
0.206
Permeability,

Loose
bed


0.025
0.037
0.013
0.093
0.283
0.012
0.015
0.045
0.130
0.238






Compact
bed


0.015
0.025
0.005
0.069
0.238
0.008
0.012
0.089
0.106
0.193





, ^

Mean
P


0.020
0.031
0.009
0.081
0.261
0.010
0.014
0.067
0.118
0.216
0.200
0.053
0.027
0.014
0.009
Gravel

ID



1
2
3
4
5
A
B
C
D
E
14
15
16
17
18
Sample1

Symbol
used in
Figures
4 and 5

e




A




Q




1 based on data from Cooper (1965);  ID column allows data in this table to be related to that of Cooper.

-------
    .40
    .35
    .30
    .25
2   -20-
    .15
    .10
    .05
                                        e = .176 + .028d
                                           (r  = .85)
                                    NOTE:
                                      Data   taken from Table 2 and
                                         based on Cooper(1965)
                GEOMETRIC MEAN PARTICLE  DIAMETER ,dg, centimeters
       Figure 4.   Relationship  between sediment porosity  and
                    geometric mean sediment particle  diameter.

                                    13

-------
    .28
              NOTE:
               Data   taken from Table 2 and
                  based on Cooper(1965)
e
HH
—I
I—I
3
    .24
    .20
     12
    .08
    .04
                                                          1.92
                   GEOMETRIC MEAN PARTICLE DIAMETER, d , centimeters
   Figure 5.   Relationship  between  gravel  permeability p,  and
                geometric mean sediment particle diameter.

                                    14

-------
obvious if  we recall the  definition of  d16,  i.e., the  size below which  16
percent of  the gravel  is  finer.   Naturally  we should not always expect  that
6.3 mm and d16 coincide as it did in this case.


 TABLE 3.   COEFFICIENT OF DETERMINATION BETWEEN PERMEABILITY  AND POROSITY  AND

           INDICES OF GRAVEL COMPOSITION d ,  ZJj AND PERCENT FINES USING

           LINEAR AND POWER FITTING FUNCTIONS

Sediment
property
Porosity, e
Permeability, 3
d.
9
.85
.90
Z^ Percent fines
.83 mm
.71 .79
.97 .82
less than
6. 3 mm
.77
.93

     In  summary,  the  geometric mean  diameter is  recommended as  a  standard
measure  for substrate  characterization  in fisheries  work for the following
reasons:

     (1)  d   is  a  conventional  statistical   measure  being  used  by  several
          disciplines to represent sediment composition.

     (2)  d   is  a  convenient  standard  measure  that  enables comparison  of
          sediment sample results between two studies.

     (3)  d  is calculated  from d84 and d16,  two  parameters  that can be used
          to calculate the standard deviation.

     (4)  d  relates to the permeability and porosity of channel sediments and
          to embryo survival,  at least as well  as "percent fines".

     (5)  d  is a more complete description of total sediment composition than
          "percent  fines" and sediment composition  evaluations  in many cases
          involve less sampling  error using d .

     (6)  Because d  relates to porosity and permeability, it is potentially a
          suitable ^unifying  measure   of  channel  substrate  condition  as  it
          impacts embryo survival.
                                       15

-------
                                  PROCEDURES

     Three  investigators  determined  the  sediment  composition  of  selected
spawning  areas  from 1966 to 1977 in  the  Salmon River drainage.  They used at
least  three different  procedures  in site  selection,  method  of collection,
equipment and analysis.           '

     The  data  collected by Ortmann (1968) for 1966 and Platts (1968, 1970 and
1972) for 1967-1974 were obtained using the McNeil method with 6-inch (153 mm)
diameter  cores.   The USDA  Forest  Service Materials  Testing Laboratory,  Salt
Lake City,  Utah,  heat-dried,  screened, and weighed the selected particle size
groups  of  the samples  collected  by  Platts.   Platts  collected cores  along
permanent  stratified random transects  crossing spawning areas.  Two  samples
were taken, one each at  1/4 and 3/4 intervals across  each transect.   Occasion-
ally a third sample was taken mid-point on  the transect.

     Corley (1975a,  1975b  and  1978) collected samples from  1975  through 1977
with a  12-inch (305 mm) core sampler.  About 5 gallons (18.9 liters) of sedi-
ment  was collected  with  each  sample.   The  sediment samples  of  Ortman  and
Corley  were  sieved wet  and analyzed  in  the  field  using  standard  sorting
screens  for sediment separation.  Weights of the selected sediment size groups
were determined  using the  volumetric water displacement method  suggested by
McNeil (1964),

     Corley selected gravel areas  from 25 x  25  foot (7.63  x  7.63 m) square
grids laid  out within  known spawning riffles.   Ten core samples were selected
randomly  within  the designated  625  ft2  (58.1  m2)  square.   Four  riffles,  all
located  within the  spawning area,  were selected  to represent  the complete
spawning site.

     Very fine particle sizes,  on the order of .0025 in (63 microns) and less,
were analyzed by Corley using an Imhoff cone and by Platts using a hydrometer.
The mass  fraction of  these small  particles per  sample was  much less than 1
percent.

     The  treatment  of  very  large  sediment particles  was more difficult  and
depended  on  the core  diameters  used.  Frequently  large  particles  were found
obstructing the 6-in (153  mm)  core sampler, in which case they were added to
the sample.  The  use of a  12-in (305 mm) core sampler may  present a  smaller
sampling  bias.  Since  the  total  volume  from a 6-in core  sampler  is  smaller
than that taken from a 12-in core sample,  the  presence  of large particles in
the small sample  could skew the distribution,  biasing it toward coarse compo-
sition.   This  would cause  larger  fluctuations  in  the results.   Also,  in  the
process of  digging  out  the channel  materials  within the  core sampler,  fine
sediments are more readily collected than large siz«  particles.
                                      16

-------
     Data obtained by  Platts  were presented for 3-in (76.2 mm) size particles
and  less,  i.e.,  that  fraction  of the  sample  passing  the 3-in  sieve.   All
materials above 3 in were grouped into one size class.  Corley's 1975 and 1977
data are  analyzed only  for  sediment  particles 1-in  (25.4  mm)  and less.   The
composition above  1 in and below 3 in was not sorted, except for  1976 data.
                                       17

-------
                           RESULTS AND CONCLUSIONS

      Data  describing  the particle  size  distributions  for  various  spawning
 areas are  summarized  in  Tables  4-7.  Because of the differences in screen size
 selection  by the different authors, the tables  do not show directly-measured
 data  for all sieve sizes.  Instead, interpolated  values  (shown within paren-
 theses)  are  inserted  for convenience.   The interpolations were made by graph-
 ing  the  particle size frequency distribution curve for each sample and taking
 the interpolated number from its respective place on the curve.

      The substrate compositions of Chinook spawning areas located in the South
 Fork  Salmon  River,  Middle Fork Salmon River tributaries, and the Salmon River
 and  one  of  its  tributaries are shown  in Tables 4 and 5.   Each spawning area
 listed  represents  from  5  to  130  core  samples collected  from that site.   In
 addition,  averages  for all  chinook spawning areas  located  within each of the
 three river  drainages are presented,  and, finally,  a grand  average  for  all
 chinook  salmon  spawning areas  is given.  The preference  for  the  sediment
 composition  chosen  by spawning salmon  is  reflected by the  d  averages of .28
 to  .79 in (7 to 20 mm), depending on the river  reaches  sampled.  The narrow
 range salmon find acceptable  for spawning becomes apparent  when the  average
 sediment particle size found  in spawning areas is compared with other channel
 reaches  of  similar  size they  could  have selected  for spawning,  e.g.,  they
 could have selected  fine  sand with d   less than  0.04 in (1  mm)  or areas of
 predominant rubble with d greater than%.99 in  (100 mm).

      Orcutt et al.  (1968) listed the preferred substrate size used by spawning
 steel head troutTn Idaho  as between .25  in (6.7 mm) and 4.0 in (101.6 mm).

      Based on the 815 samples taken from the  12 most important salmon spawning
 areas  in Idaho,  channels used for spawning averaged only 8 percent fine sedi-
 ments below  .03  in  (.83 mm)  in particle size.   However,  these areas averaged
 30 percent in sediment  particle  size  less than .19  in  (4.7  mm).   This indi-
 cates that entrapment of alevins by fine sediments may be more of a problem in
 the Salmon River drainage than embryo  or  alevin mortality  caused by low dis-
 solved oxygen in the  subsurface flows.   About 93 percent of the sediments  are
 less  than  3  in  (76.1  mm) in  particle  diameter,  which  shows  salmon  are  not
 looking  for  large  sediments  for spawning.    Actually,  the  majority   of  the
 sediments they are using  are less than .75 in (19 mm) in particle size.

     There are differences in sediment  sizes  used for spawning between  streams
or areas within  streams, but  these are  not major  differences.  The change in
procedures  from  year  to  year  and person  to  person  may  have  some  effect on
these differences.

     In comparing particle sizes used by salmon between the three major drain-
ages, the differences were again not substantial.  There was a difference of 4

                                      18

-------
 TABLE 4.  CHANNEL SUBSTRATE COMPOSITION BY YEARLY AVERAGES BY SEDIMENT PARTICLE SIZE IN SAMPLES TAKEN FROM CHINOOK  SALMON SPAWNING AREAS.
Sampl e
Stream or Area Size
Time
Period
Substrate Particle Size by Groups Representing Percent Volume Passing1
Through Sieve of the Designated Size (mm)
76.1
50.8
38.1
25.4
19.0 12.7
9.51 6.35
4.76
2.83 2.38
2.00 1.00 .83
.42 .25 .21
.10 .07 .05
South Fork Salmon River
Stolle Meadows
Area
Dollar Area
Poverty Area
Oxbow Area
Glory Area
Johnson Creek
145
40
310
50
80
100
1966
to 1975
1975
1966
to 1976
1975
1966
to 1975
1966
to 1976
92
—
93
--
93
86
79
--
84
—
82
74
71
—
77
—
70
(63)
53
49
68
74
57
51
48 42
(44) (40)
60 52
(68) (63)
52 46
(44) 38
38 34
(35) 31
47 42
(58) 53
41 36
(33) 28
30
28
37
48
31
25
26 (23)
(23) (19)
33 28
(41) (33)
(28) (24)
21 17
19 15 11
(14) (10) 5
23 18 14
(25) (17) 9
20 (15) 12
(14) 10 8
6 2 1
(4) (2) .5
742
(7) (4) 1
5 2 1
(5) (3) 1
.6 .1 0
(.5) (.4) 0.4
1 0 0
(1) (0) 0
1 .5 .5
(1) (0) 0
Middle Fork Salmon River
Bear Valley Creek
Elk Creek
Loon Creek
Salmon River
Lower Decker Area
Upper Decker Area
Alturas Creek
Combined Average
Total Sample Size 815
20
20
20
5
5
20


1968
1968
1969
1969
1969
1969


98
100
94
82
96
95
93

90
98
76
62
85
77
81

82
89
67
56
78
66
72

72
71
55
49
67
54
60

63 53
58 45
47 38
43 37
59 49
46 38
53 45

48 (42)
39 (34)
32 (27)
32 (26)
42 (35)
33 (27)
40 34

37
28
22
21
28
24
30

(33) (28)
(24) (19)
(20) (17)
(18) (15)
(23) (19)
(22) (17)
26 22

24 (18) 12
15 (10) 5
15 (11) 7
12 (9) 6
14 (10) 5
14 (10) 6
17 13 8

5 1 (1)
3 .5 (.5)
2 1 (1)
3 1 (1)
1 0 0
3 1 (1)
4 2 0.9

000
000
000
000
000
000
0.4 0.1 0.1


1 Values in parentheses are interpolated by graphing the particle size distribution curve and selecting the percent passing from the
  intersection of the group size with the curve.

-------
    TABLE 5.  CHANNEL SUBSTRATE COMPOSITION BY DRAINAGE BY SEDIMENT PARTICLE SIZE IN SAMPLES TAKEN FROM CHINOOK SALMON SPAWNING AREAS.

Substrate Particle Size by Groups Representing Percent Volume
Through Sieve of the Designated Size (mm)
Drainage
South Fork
Salmon R.
Middle Fork
Salmon R.
Main Salmon
River
Sample
Size
725
60
30
76.1 50.8 38.1 25.4 19.0 12.7 9.51
91 80 70 59 53 47 42
97 88 79 66 56 45 40
91 75 67 57 49 41 36
6.35
37
(34)1
(29)
4.76
33
29
24
2.83
29
(26)
(21)
2.28
24
(21)
(17)
2.00
19
18
13
1.00 .83 .
14 10
(13) 8
(10) 6
Pass'ing1
42 .25 .21 .10
63 1 0.8
3 0.8 (0.8) 0
2 0.7 (0.7) 0

.07 .05
0.2 0.2
0 0
0 0
Values in parentheses are interpolated by graphing the particle size distribution curve and selecting the percent passing from
the intersection of the group size with the curve.

-------
percent at .03  in  (.83  mm) and less  in  particle  size,  9 percent at the  .19 in
(4.7 mm)  particle  size, and  6 percent at the 3 in (76.1 mm) particle size.
Salmon are not  searching out  major differences in sediment particle sizes for
spawning regardless of  the  drainage,  stream  or  stream area.   However, studies
have shown  that an increase  from 5  percent  to  15 percent in fine sediments
less than  .03 in  (.83  mm) in particle size  can result  in a change from low
mortality to  high  mortality.   Therefore,  salmon  have to search out sediments
within narrow particle  size distribution  limits  because the  survival require-
ments of the embryo and alevin  are  so demanding.

     In an attempt to  find a  best correlation between various definitions of
"percent fines" and geometric  mean diameter  for  chinook spawning  substrate in
the Salmon River, a power curve fitting procedure  of the form

                         (percent < d) = A(d )b
                                           y
was  used,  where d is the  appropriate particle  diameter below which the  mass
fraction percentile  is  finer,  and A and  b  are  constants.   This formula was
repeatedly applied to the  core data  and the  coefficient  of  determination, r2
was calculated.

     "Percent fines"  less than .08 in (2 mm)  provides the  best fit (Figure 6).
Possibly  this  is  because   .08  in  (2 mm)  coincides  with  the  mean  of  the  16
percent!les of the entire data sample for the spawning substrate  in  the  Salmon
River drainage.

     The curves in Figure  6 might be used as a summary and as rough estimates
of  "percent fines" in spawning areas  in the Salmon River drainage.  The  figure
provides  a  good illustration  of  the value  of   d   in synthesizing  apparently
unrelated results.  A vertical  line  drawn througfi a  d of .24 in (6 mm), for
example,  shows  that  this  d  is equivalent to each  ofgthe following "percent
fines"  specifications:   22 percent  less  than .04 in  (1  mm),  26  percent less
than  .08 in  (2 mm),  31 percent  less than .09   in  (2.38  mm),  36  percent less
than  .11  in  (2.83 mm),  and 39  percent  less  than .19  in  (4.76 mm).   It should
be  emphasized,  however, that  use of  Figure  6  is  restricted to  obtaining an
estimate of fine sediments  for  this  specific  data set.  Figure 6 should not be
used to determine  a general  relationship for other spawning substrates.

     The average  substrate composition for spawning  areas located  in  each of
the three drainages listed  in Table 4 is plotted  using semi-log axes in Figure
7.   The  consistently  coarser  structure of substrates  used  by spawning  salmon
in  the  Salmon River  and its tributary, Alturas  Creek, relative to those areas
used  in  its  two major  tributaries, the Middle Fork Salmon River and the South
Fork  Salmon  River, is  clearly shown.  The upper Salmon  River as well  as the
Middle  Fork  Salmon River are used mainly  by  spring chinook  salmon.  There is
some  indication that  spring chinook  salmon spawning areas in Idaho consist of
a  coarser  substrate.   However,  the data alone cannot  be  used to substantiate
this because  the South  Fork Salmon River (used by summer chinook) may still be
affected by past logging operations.
                                       Zl

-------
                                                    NOTE:
                                                      l.Data for these curves are the
                                                        yearly averages used 1n
                                                        calculating the longer term
                                                        averages of Table 4.
                                                      2.Data points  shown for d = 2.00mm
                                                        only.
                                          = 4.76mm (r =  .79)

                                       -d  = 2.83nm (r2= .82)

                                       •d  = 2.38mm (r2=  .84)

                                       -d  = 2.00mm (r2=  .86)


                                        •d  = 1.00mm (r2=  .85)
                            GEOMETRIC MEAN DIAMETER, dQ, centimeters
Figure  6.   Relationship between geometric mean  diameter  and  percent  fines.
                                            22

-------
ro
CO
                      GO

                      o LU


                      CO CO


                      Q. LU
l-l CO
a ui
LU a
    700



     90



     80



     70



     60



S    50



     40



     30



     20



     10
                                                           South Fork Salmon River
                                             Middle  Fork Salmon River & Tributaries
                                          Main Salmon River & Tributary
                                                                                                    J	1	1	L.
                                                                                             10.
                                                                                                               ' ' ' .

                                                                                                                 100.
                                                             SIEVE SIZE, millimeters
                          Figure  7.   Particle size distributions for  Chinook salmon  spawning areas.

-------
     It  is difficult  to provide  a  systematic  comparison  showing  substrate
composition differences between the streams, spawning areas and stream reaches
sampled.  The main  difficulty  arises  because of the change  in  procedures  and
equipment  from year  to  year  and  person to  person.   Therefore,  some of  the
variability  and differences  indicated  by  the  data  might  be  procedural  in
nature  and  not  a  reflection of the true situation.  Taking into consideration
this problem,  an  attempt  is  made to  compare  these  variabilities.   Thus,  the
data sets are selected to avoid some of the more obvious problems.

     A  detailed look at  a  single  spawning  area (Poverty area)  on  the  South
Fork Salmon River  is given  in Table 6.   Four sites  were  sampled  and  five
samples were taken  at each site.   The geometric mean particle diameter and to
some extent "percent  fines"  within each site  as well  as  between sites varies
by  a  two-fold   magnitude.  Thus,  there is some  variability  between  each  site
within  the  spawning area.   This was expected  as the  upper end  of the Poverty
spawning  area   is  composed mainly  of  rubble with  gravel in  the  downstream
direction to gravel mixed with fine sediments at the lower end of the spawning
area.


 TABLE 6.  VARIATION OF GEOMETRIC  MEAN PARTICLE DIAMETER OF SPAWNING SEDIMENTS
           IN THE  POVERTY AREA AMONG SAMPLES COLLECTED IN 1976.
                                Percent fines                  d  mm
                                   <6.3 mm                      9
 Site          Point                       Average                    Average
1 1
2
3
4
5
6.7
7.7
6.7
8.3
9.0
8.5
11.7
7.8
8.5
6.9
                                             7.7                        8.4

   2        All points         10,6                        6.4
   3            "               8.3                        8.1
   4            "              10.8                       12.0
                                             9.4                        8.4
     Particle  size  distributions  for  sediment collected  from areas  used by
trout  spawning  and  rearing  are  listed  in  Table  7.  These  sample  areas were
distributed over  a  much  larger  portion of  the stream channel  than were the
samples collected in  the  salmon  spawning areas discussed earlier.  Each hori-
zontal line in  Table  7 represents the average  for  several  individual samples
taken  from each stream.   Overall  averages  for  tributaries  within  the  three
major  rivers and a  grand  average for all  three  areas are presented.  Because
of the lack of the  upper portion (i.e., particles larger than 1 in (25.4 mm))
                                      24

-------
TABLE 7.  CHANNEL SUBSTRATE COMPOSITION BY YEARLY AVERAGES BY SEDIMENT PARTICLE SIZE CLASS IN SAMPLES TAKEN FROM TROUT SPAWNING AND REARING  AREAS.
Drainage Sample
Stream Size
South Fork Boise River
Fall Creek
E.F. Fall Creek
W.F. Fall Creek
Bear Hole Creek
Trinity Creek
Spring Creek
Spring Creek
Johnson Fork
Steel Creek
North Fork Boise River
N.F. Boise River
Payette River
Squaw Creek
Second Fk. Creek
Third Fk. Creek
70
10
5
5
15
15
5
5
10
45
40
15
10
15
Combined Average by Stream
Total Sample Size 155

Substrate
Year 76. 1
1975
1975
1975
1975
1975
1975
1975
1975
1975
1976 81
1974 60
1974 54
1974 59
1974 66
65

50.8 38.1 25.4
44
40
32
50
58
35
55
37
36
(72) (63) 54
(53) (46) (40)
(47) (40) (34)
(51) (43) (36)
(60) (54) (48)
(58) (50) 43

19.0
(39)
(37)
(28)
(47)
(53)
(30)
(49)
(33)
(33)
(48)
(33)
(27)
(28)
(42)
(38)

Particle Size by Groups
Sieve of the
12.7
(35)
(34)
(24)
(43)
(47)
(26)
(44)
(29)
(29)
43
27
21
21
36
33

9.51
(31)
(31)
(20)
(40)
(41)
(22)
(39)
(25)
(25)
(37)
(25)
(19)
(19)
(34)
(29)

6.35
27
28
16
36
35
18
34
21
21
32
(22)
(16)
(17)
(31)
25

4.76
24
26
14
33
32
16
30
19
18
29
19
13
15
28
23

Representing
Designated
2.83
(20)
(22)
(12)
(28)
(27)
(13)
(25)
(15)
(15)
(24)
(18)
(11)
(13)
(26)
(19)

2.38
(17)
(19)
(10)
(23)
(23)
(ID
(20)
(12)
(12)
19
(17)
(10)
(12)
(24)
16

Percent Volume Passing1
Size (mm)
2.00
(13)
(15)
(8)
(18)
(18)
(8)
(15)
(9)
(9)
(15)
16
8
10
22
13

1.00
(10)
(12)
(6)
(13)
(14)
(5)
(10)
(6)
(6)
(11)
(14)
(6)
(9)
(18)
10

.83
6
8
3
8
9
2
5
3
4
7
(12)
(5)
(7)
(15)
6

.42
(5)
(7)
(2)
(7)
(7)
(2)
(4)
(3)
(3)
(5)
(9)
(4)
(6)
(12)
(5)

Through
.25
(3)
(5)
(1)
(5)
(5)
(1)
(3)
(2)
(2)
(3)
(7)
(3)
(4)
(9)
(4)

.21
1
3
0
3
3
0
2
1
1
1
(4)
(2)
(3)
(6)
2

.10
(1)
(3)

(3)
(3)

(2)
(1)
(1)
(1)
(2)
(1)
(1)
(3)
(2)

.07 .05
(1) 1
(2) 2

(2) 2
(2) 2

(1) 1
(0) 0
(0) 0
(0) 0
0
0
0
0
0.5 0.6

1 Values in parentheses are interpolated by graphing the particle size distribution curve and selecting the percent passing from the intersection
  of the group size with the curve.

-------
 of  particle size distribution, geometric mean diameters for trout areas could
 not be calculated.   However,  to provide a  comparison,  the particle size dis-
 tributions  averaged  for all chinook spawning areas  vs.  all samples collected
 in  trout  channels  used  for  rearing and  possible  spawning are  presented in
 Figure 8.   The coarser substrate in resident trout channels is clearly shown.
 Trout  often  spawn  in  small  niches within  the  channel that  frequently have
 finer  substrate than the overall riffle  areas.   Therefore, redds are usually
 interspersed  among areas  of  much  coarser material.   The sampling procedure
 does not take this  into  account.

     Table  8  shows  the  variation  of  geometric mean  diameter  for  1976 for
 different  spawning areas  in  the  South Fork  Salmon River.  These  areas are
 arranged in order of increasing channel elevation, showing  that fish have used
 persistently  coarser  spawning  gravel  in the upstream direction, with d  = .58
 in  (14.7 mm)  in the upstream reaches compared with d  = .35 in (8.8 mm)9in the
 downstream  reaches of  the South Fork Salmon River. 9Whether this is a reflec-
 tion of availability or  preference for certain sediments  is  not determined.

     An  attempt was  made  to  obtain  measurements within  egg pockets  in the
 Poverty  area.  Fifteen  freeze core  samples were  collected during  the 1977
 spawning period.  The  freeze  core  rods were driven to a depth of 18 inches in
 the  substrate.  Results  were  analyzed  separately for each  core  sample  and in
 combination.  The geometric mean diameter for the combined  15 core samples was
 18.4 mm.  The geometric mean diameter for top to 6 inches, 6 to 12 inches, and
 12  to  18  inches were  respectively 20.3 mm, 22.4 mm,  6.5 mm.   Unfortunately,
 the  individual analysis  of separate core  samples  revealed  unusual  scatter.
 For  example,  the average  d  for the  15  samples was 34.6  mm with  a standard
 deviation of 20,6 mm from thvs mean.
           TABLE 8.  VARIATION OF GEOMETRIC MEAN DIAMETER OF SPAWNING
                     SEDIMENTS IN SAMPLES COLLECTED FROM CHANNEL REACHES
                     IN THE SOUTH FORK SALMON RIVER IN 1976
         Site                                       d (mm)
                                                     y

Downstream reaches:

     Glory Hole                     9.6
     Oxbow Area                     8.5            Average =             8.8
     Poverty Flat Area              8.4


Upstream reaches:

     Dollar Creek Area             13.5            A                    ,, ,
     Stolle Meadow Area            15.8            Average =            14.7
                                      26

-------
in to


3 o
«-i LU
s D;
a uj
LU o
in
100





 90





 80





 70





 60





 50





 40





 30





 20





 10
                                               Chinook Spawning Areas
                                      Resident Trout Channels
                                                1.                  10.




                                           SIEVE SIZE,  millimeters
        Figure 8.   Comparison of particle size distributions  for resident

                     trout channels and Chinook salmon  spawning areas.

-------
     Continued attempts  to  characterize the strata of egg pockets resulted in
Table 9 which was obtained  from the Poverty area in November 1978.  The sample
was  collected  with  a battery of  freeze cores  and an attempt was  made to ex-
tract an  entire redd  egg pocket soon  after spawning.  The  dry sample weight
was  620 kg  and the  geometric mean diameter  of the redd was 23.3 mm.  This is
somewhat large compared with similar measurements taken by other means both in
1978 and previously.

     The difference is attributed to two important factors.  The first is that
considerable coarsening  of  the  gravel  was accomplished  by the  fish during
spawning.    The  fish was observed digging deep,  and  covering  the  eggs  with
relatively  coarse  substrate.   The digging action  released considerable fines
thus  rendering the  texture relatively  coarse compared with the  surrounding
gravel.   The second explanation  lies  in the bias introduced in wet seiving of
the 1975 through 1978 gravel samples,  even though they were obtaind with 12-in
core, which  is probably  an  adequate sample size.  In  this process,  the water
held within the  space between  small  particles  is artificially added to the
size fraction.  The  larger particles do not hold much excess  water and thus
are  relatively unaffected  by wet  seiving.    The  method   is therefore unduly
biased toward  smaller  particles.  An estimate was made of this bias  on 1977
data in the Poverty  area.   The  average of 40 samples gave d  = 8.4 mm without
correction.   With an approximate  correction, d  = 11.9 mm, °.e.,  a bias of 42
percent.                                        9

     Composition of the  egg pocket  in the vertical  shows  that  the d  for top
to 6 inches, for 6  to  12 inches and  12  to 18 inches  are respectively939.2 mm,
20.1  mm, and 35.2 mm.

          TABLE 9.   ANALYSIS OF  A COMPLETE CHINOOK SALMON  EGG POCKET
                    TAKEN IN POVERTY AREA DURING 1978 SPAWNING
               Particle size (mm)                Percent fines
203.2
152.4
127.0
101.6
76.2
50.8
25.4
12.5
6.3
4.75
.84
.246
.074
.074
100.0
98.9
93.4
82.2
62.4
50.6
37.4
25.4
18.48
14.98
2.78
0.48
0.08
0
                                      28

-------
                                  DISCUSSION

     While there appears to be a slight difference in the procedures discussed
for obtaining samples and presenting data on the description and evaluation of
spawning  habitat for  salmonids,  these  are minor  indeed,  compared with  our
inability to relate these procedures to the effects created by different types
of land  use,  an area of impact evaluation  that begs for a better understand-
ing.   That goal  will  be achieved  by establishing more unified, scientifically
defensible procedures.   The  authors know  of no place in  Idaho,  for example,
where the  effects of  a land  use  such as logging and road construction have
been accurately related to  the  reproductive success  of a  chinook salmon or
steel head  trout population.    For  proper  land   use  and  fishery  planning  and
management, this degree of predictability should be attained.
                                      29

-------
                                  REFERENCES

Bogardl,  J.   1974.   Sediment Transport in Alluvial Streams.  Akademiai  Kiato
     Budapest.  826 pp.

Cooper, A. C.  1965.  The  effects  of transported stream sediments on the sur-
     vival of sockeye and  pink salmon eggs and alevin.  International Pacific
     Salmon Fisheries Commission, New Westminster, B.C.   Canada.

Corley, Donald R.  1975a.  Stream inventory survey of Squaw Creek, Second Fork
     and Third Fork.  USDA For.  Serv.  Intermt. Reg., Boise Natl. For., Boise,
     ID.  59 pp.

          .  1975b.   Stream inventory  survey  of streams in the Fall Creek and
     Trinity Creek  drainages,  1975.  USDA  For.   Serv.  Intermt.  Reg.,  Boise
     Natl. For.,  Boise,  ID.  53pp.

	.  1978.   Fishery habitat survey  of the South  Fork  Salmon River -
     1977.  USDA  For. Serv.,  Intermt.  Reg.,  Boise  Natl.  For.,  Boise, ID.  90
     pp.

Hall,  J.  D. ,  and  R.  L. Lantz.   1969.  Effects of  logging  on  the habitat of
     coho  salmon and  cutthroat trout in  coastal streams,  jji Symposium on
     salmon  and  trout  in  streams,  T.  G.  Northcote,  Ed.,   Univ.  British
     Columbia,  Vancouver, B.C.,  pp.  355-375.

Innman, D.  L.   1952.  Measures for  describing  the size distribution of  sedi-
     ments.  Journal  of  Sedimentary  Petrology, 22(3):125-145.

Iwamoto,  R.  N.,   E.  0.  Salo,  M. A.  Madej  and R.  L. McComas.   1978.  Sediment
     and water quality:  A review of  the  literature including  a suggested ap-
     proach for water quality criteria.  EPA 910/9-78-048, 1978.

McNeil, W. J.  1964.  A method of  measuring  mortality  of  pink  salmon eggs and
     larvae. U.S. Fish  and Wildl. Serv. Fish. Bull.  63(3):575-588.

McNeil, William  J.,  and W. H.  Ahnell.   1964.   Success  of  pink  salmon spawning
     relative to  size  of  spawning  bed materials.   U.S.  Fish  and Wildl.  Serv.,
     Spec. Sci.  Rep.  Fish., 469, 15 p.

Ortmann,  David W.   1968.   Particle size of  substrate materials in  tributaries
     and  the  South  Fork Salmon River.   Idaho Fish &  Game  Dep.   Addendum to
     F-49-R-5, Job. No.  1,  Amendment No. 1, Boise, ID, 4 p.

Orcutt, D.  R., B.  R.  Pulliam  and Arthur Arp.   1968.  Characteristics  of  steel-
     head trout redds in Idaho streams.  Trans. Amer. Fish.  Soc.  97(l):42-45.

                                        31

-------
Platts,  William  S.   1968.   South Fork  Salmon River,  Idaho,  aquatic habitat
     survey  with  evaluation of  sediment  accurement,  movement  and damages.
     USDA For. Serv.  Intermt.  Reg., Ogden, UT, 137 p.

	.   1970.  The effects  of logging and road construction on  the aqua-
     tic habitat  of the South Fork Salmon  River,  Idaho.  Proc. Fiftieth Annual
     Conf.  of the Western Assoc. of State Game and  Fish Comm., p. 182-185.

	.   1972.  Aquatic  environment and  fishery study  South Fork Salmon
     River,  Idaho,  with evaluation  of sediment  influences.   USDA For.  Serv.
     Intermt. Reg., Ogden,  UT, 106 p.

Platts,  William  S.  and Walter  F.  Megahan.   1975.   Time  trends  in riverbed
     sediment composition  in salmon  and steel head spawning areas:  South Fork
     Salmon  River,  Idaho.   Trans, of the 40th North  Am.  and Natl.  Resources
     Conf., Wildl. Manage.  Inst., Washington, D.C.  pp.  229-239.

Vanoni,  V.  A.  1977.   Sedimentation  Engineering.  American  Society  of Civil
     Engineers, 345 East 47th St. , New York.

Walkotten,   William  J.   1976.   An  improved  technique  for  freeze  sampling
     streambed  sediments.  USDA  For.  Serv.  Res.   Note PNW-281, 11 pp.  Pacific
     Northwest For.  and Range Exp. Stn., Portland,  Oreg.
                                       32

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
  REPORT NO.
        EPA-600/3-79-043
                                                          3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
       Sediment Particle Sizes  Used  by Salmon for
       Spawning with  Methods for Evaluation
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

       William S. Platts, Mostafa  A.  Shirazi  and
       Donald H. Lewis	
                                                          8. PERFORMING ORGANIZATION REPORT NO.
j PERFORMING ORGANIZATION NAME AND ADDRESS
 [nv.Res.Lab.-Corvallis   and   Intmtn  For.  and Range ExSta
 )ff.of Res. and Dev.          Forest  Service
Envirn. Prot. Agency          U.S.  Dept of Agriculture
Corvallis, OR 97330           Ogden,  Utah  84401
                                                           10. PROGRAM ELEMENT NO.
            11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
       Environmental Research  Laboratory--Corvallis
       Office of  Research and  Development
       Environmental Protection Agency
       Corvallis, Oregon  97330	
                                                           13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE
              EPA/600/02
15. SUPPLEMENTARY NOTES
       Contact: Mostafa A.  Shirazi,  Corvallis, OR 97330 503-757-4751  (FTS  420-4751)
  ize composition of  substrates  used by chinook salmon for spawning  in the South Fork
 Salmon River, the main  Salmon River and tributaries of the Middle Fork Salmon River, ID
 was determined. Substrates  used by resident trout were analyzed  for streams in the Bois
 and Payette River drainages. These analyses were made over time  and space to determine
 particle sizes preferred  by spawning salmon, yearly differences  in  sizes used by these
 salmon, the size differences used by spring and summer chinook salmon, and differences
 between channel sediments used  by chinook salmon for spawning and those substrates
 occupied by trout. Use  of the geometric mean particle diameter method is presented as a
 companion measurement to  "percent fines" for an easier and more  complete analysis of
 sediments used for spawning.  The geometric mean particle diameter  is more adaptive to
 statistical analysis than the more common method of using "percent  fines." The geometri
 mean diameter of the sediment particle size distribution is  used for analyzing channel
 sediments and its relationship  to "percent fines," substrate permeability, and substrat
 porosity is established.  For sediments used by spawning salmon, the strongest correla-
 tion between the two methods of analysis, "percent fines" and geometric mean diameter,
 was for fine sediments  below 0.08 in (2mm) in particle size.  Chinook salmon selected
 sediments for spawning  that were mainly between .28 and  .79  in  (.7  to 20 mm) in geome-
 tric mean particle diameter, regardless of stream selected.  The composition of spawn-
 ing sediments selected  by chinook salmon each year between 1966-76  were quite uniform.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                          c.  COSATI Field/Group
  substrate
  sediments
  salmon
  spawning
  gravelbed  stream
sediment particle  sizes
ISTDISTRIBUTION STATEMENT

 Release to  public
19. SECURITY CLASS (ThisReport)

 unclassified	
21. NO. OF PAGES
                                              20. SECURITY CLASS (This page)
                                               unclassified
                           22. PRICE
 EPA Form 2220-1 (Rev. 4-77)
                                             33

-------