&EPA
            United States
            Environmental Protection
            Agency
          F iivironment.il Research
          Laboratory
          Athens GA 30613
EPA 600 3 82-024.
May 1982
            Research and Development
Methodology for
Overland and Instream
Migration and  Risk
Assessment of
Pesticides

-------
                                        EPA-600/3-82-024
                                        May 1982
 METHODOLOGY FOR OVERLAND AND INSTREAM MIGRATION
        AND RISK ASSESSMENT OF PESTICIDES
                         by

Y. Onishi, S.M. Brown, A.R.  Olsen, M.A.  Parkhurst,
            S.E. Wise, and W.H.  Walters

                      Battelle
           Pacific Northwest Laboratories
             Richland, Washington 99352
              Contract No. 68-03-2613
                  Project Officer

                 Robert B. Ambrose
   Technology Development and Applications Branch
        Environmental Research Laboratory
             Athens, Georgia  30613
        ENVIRONMENTAL RESEARCH LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
             ATHENS,  GEORGIA  30613

-------
                                 DISCLAIMER

     This report has been reviewed by the Environmental Research Laboratory,
U.S. Environmental Protection Agency, Athens, Georgia, and approved for pub-
lication.  Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
                                      11

-------
                                 FOREWORD

      As environmental controls become more costly to implement and the
penalties of judgment errors become more severe, environmental  quality
management requires more efficient management tools based on greater know-
ledge of the environmental phenomena to be managed.  As part of this
Laboratory's research on the occurrence, movement, transformation,  impact,
and control of environmental contaminants, the Technology Development  and
Applications Branch develops management and engineering tools to help pollu-
tion control officials achieve water quality goals through watershed manage-
ment.

      Many toxic contaminants are persistent and undergo complex interactions
in the environment.  As an aid to environmental decision-makers, the modeling
technique described in this report was developed to simulate overland and
instream transport of toxic contaminants and to predict the probability of
acute and chronic impacts on aquatic biota.

                                      David W. Duttweiler
                                      Director
                                      Environmental Research Laboratory
                                      Athens, Georgia

-------
                                 ABSTRACT

      To provide planners and decision makers in government and industry
with a sound basis for environmental  decision making,  the Chemical  Migration
and Risk Assessment (CMRA) Methodology was developed to predict the occur-
rence and duration of pesticide concentrations in surface waters receiving
runoff from agricultural lands and to assess potential  acute and chronic
damages to aquatic biota.  The CMRA methodology consists of overland pesti-
cide transport modeling, instream pesticide transport  modeling, statistical
analysis of instream pesticide concentrations, and risk assessment.

      The CMRA Methodology uses two state-of-the-art overland and instream
models, ARM and SERATRA, to continuously simulate nonpoint source pollution
processes.  It is useful for evaluating both short- and long-term migration
and fate of both dissolved and particulate pesticides.   The risk assessment
procedure, coupled with statistical analysis of predicted pesticide concen-
trations by the computer program FRANCO and pesticide  toxicity data, provides
a good scientific basis for pesticide risk assessment.   Because of a lack of
extensive knowledge on pesticide toxicity, however, the risk assessment pro-
cedure includes only the direct effects of dissolved pesticide concentrations
on aquatic biota.  The risk assessment does not include ingestion effects or
any indirect effects such as bioconcentration or biomagnification.   In addi-
tion to pesticides, the methodology is general enough  to handle heavy metals,
radionuclides, and other toxic contaminants.

      To examine its applicability and limitations, the CMRA Methodology was
applied to the Four Mile Creek Watershed (Iowa), including Four Mile Creek
and Wolf Creek, to evaluate migration and fate of the  pesticide alachlor,
and its potential acute and chronic impacts on different fish species.  A
risk assessment was also performed for the pesticide toxaphene.  The study
demonstrated that the CMRA Methodology is capable of predicting fate and
effects of pesticides for a wide range of compounds, agricultural lands,
receiving streams, and aquatic biota.  The methodology needs additional
testing, however, to confirm its validity.  One of the shortcomings of the
methodology is that it requires extensive field and laboratory data and that
it may be limited by the availability of required data.

      Aside form its direct use by government and industry to assess the
impacts of pesticide practices, this methodology will  also provide research-
ers with a way to evaluate the relative importance of  various mechanisms that
control transport and fate phenomena occurring on watersheds and in receiving
streams, as well as a way to investigate the effectiveness of alternative
land management and pesticide control policies.
                                    iv

-------
      After development of the CMRA Methodology, the feasibility of fitting
overland and instream pesticide concentrations with statistical distributions
was examined.  This task was conducted to determine whether a simpler pesti-
cide assessment methodology that relied only on the statistical characteri-
zation of precipitation, watershed, and various receiving water body char-
acteristics could be developed to replace the more detailed CMRA Methodology.
Computer results of the Four Mile Watershed case obtained by the CMRA
Methodology were used for this analysis.   Study results indicate that dis-
solved pesticide concentrations both at stream-edge and instreams can be
expressed best by a log-normal distribution among gamma, normal, and log-
normal distributions.  To develop a statistical relationship among all
parameters describing the characteristics of precipitation, watersheds,
streams, and pesticides, the computational time required for a large number
of simulations may become excessive, however.

      Companion reports to this document are Mathematical  Model SERATRA  for
Sediment and Contaminant Transport in Rivers and Its Application to Pesticide
Transport in Four Mil_e and Wolf Creeks in Iowa, User's Manual for the Instream
Sediment-Contaminant Transport Model SERATRA, User's Manual for EXPLORE-I: A
River Basin Water Quality Model (Hydrodynamic Module Only), and Frequency
Analysis of Pesticide Concentrations for Risk Assessment (FRANCO Model).

      This report was submitted in fulfillment of Contract No. 68-03-2613 by
BatteHe Pacific Northwest Laboratories under the sponsorship of the U.S.
Environmental Protection Agency.   This report covers the period April 1978
to January 1980, and work was completed as of January 1980.

-------
                             TABLE OF CONTENTS


ABSTRACT            	       iv

LIST OF FIGURES	       ix

LIST OF TABLES	xiii

ACKNOWLEDGMENT      	      xiv

1.   INTRODUCTION    	       1

2.  CONCLUSIONS AND RECOMMENDATIONS     	       3

3.  GENERAL REMARKS ON PESTICIDES  	       7

4.  CHEMICAL MIGRATION AND RISK ASSESSMENT METHODOLOGY      .     .      12

    GENERAL DESCRIPTION OF THE CMRA METHODOLOGY   ....      12

         Overland Pesticide Modeling    ......      12

         Instream Pesticide Modeling    ......      14

         Statistical Analysis      .......      15

         Risk Assessment     ........      16

    DETAILED DESCRIPTION OF CHEMICAL MIGRATION AND RISK
    ASSESSMENT METHODOLOGY   	      17

         Overland Pesticide Modeling    ......      17

         Instream Pesticide Modeling    ......      21

         Statistical Analysis      .......      23

         Risk Assessment     ........      32

5.  APPLICATION OF METHODOLOGY     	      44

    GENERAL REMARKS     	      44

    OVERLAND PESTICIDE MODELING    	      48

-------
                         TABLE OF CONTENTS  (contd)

     INSTREAM PESTICIDE MODELING   	         .       60

     STATISTICAL ANALYSIS AND RISK ASSESSMENT     ....       84

          Case Study of Alachlor	34

          Case Study of Toxaphene  .......       89

 6.  EVALUATION OF THE CMRA METHODOLOGY	103

     OVERLAND PESTICIDE MODLEING   	      103

     INSTREAM PESTICIDE MODELING   	      103

     STATISTICAL ANALYSIS AND RISK ASSESSMENT     ....      104

     MODELING FLEXIBILITY     	      105

 7.  SIMPLIFIED STATISTICAL METHODOLOGY 	      107

REFERENCES	115

APPENDIX A:  PROCEDURE FOR CMRA METHODOLOGY APPLICATION     .    .      119

APPENDIX B:  TOXICOLOGICAL PROPERTIES OF PESTICIDES     ...      129

APPENDIX C:  DISTRIBUTION COEFFICIENTS OF ORGANIC PESTICIDES
             IN AQUATIC ECOSYSTEMS 	      171
                                     vm

-------
                               LIST OF FIGURES


 1.   Chemical migration and risk  assessment  (CMRA) methodology   .     13

 2.   Example of a discrete time step time series of  pesticide
      concentrations     .........     24

 3.   Example of LC50 curve with point exceeding region      .     .     27

 4.   Example of LC50 curve with global exceedance region    .     .     27

 5.   Postulated chronic toxicity  global exceedance region   .     .     28

 6.   Example of functions used for global exceedance summary.    .     31

 7.   Procedure for assigning LC50 values for model input    .     .     35

 8.   Procedure for assigning chronic toxicity  input  data    .     .     37

 9.   Horizontal extrapolation of  the LC50 curve to origin   .     .     39

10.   Vertical extrapolation of the LC50 curve    ....     39

11.   Horizontal extension of the  LC50 curve  .....     40

12.   Combination of the LC50 curve with the MATC value      .     .     41

13a.  The MATC line converted from MATC range     ....     42

13b.  Conservative estimate of lethality using  MATC line     .     .     42

13c.  Least conservative estimate  of  lethality    ....     43

13d.  Alternative representation of lethality in unknown regions  .     43

14.   Locations of ISU sampling stations at Four Mile Creek
      Watershed	46

15.   Four Mile and Wolf Creek in  Iowa  ......     47

16.   Simulated and measured runoff for the April 23,  1976
      storm    ...........     50
                                      IX

-------
                           LIST  OF  FIGURES  (contd)

17.    Simulated and measured runoff of the May 29, 1976 storm     .    51

18.    Simulated and measured runoff for the June 13, 1976 storm   .    51

19.    Simulated and measured runoff for the April 19, 1977 storm  .    52

20.    Simulated and measured runoff for the August 15, 1977 storm    52

21.    Simulated and measured sediment mass removal rate for the
      April 23, 1976 storm	54

22.    Simulated and measured sediment mass removal rate for the
      May 29, 1976 storm	55

23.    Simulated and measured sediment mass removal rate for the
      June 13, 1976 storm     ........    55

24.    Simulated and measured sediment mass removal rate for the
      April 19, 1977 storm	56

25.    Simulated and measured sediment mass removal rate for the
      August 15, 1976 storm	56

26.    Time variation of predicted sediment loading to Four Mile
      Creek during the three year simulation period    ...    59

27.    Time variation of predicted dissolved alachlor loading to
      Four Mile Creek during the three year simulation period     .    59

28.    Time variation of predicted particulate alachlor loading to
      Four Mile Creek during the three year simulation period     .    60

29.    Time variations of simulated concentration of sand, silt,
      clay and total sediment at Four Mile Creek River
      Kilometer 2.5 .    .    .    .    .    .    .    .     .     .    64

30.    Time variations of simulated particulate alachlor concen-
      trations associated with  sand, silt, and clay, together
      with mean particulate alachlor concentrations at Four
      Mile Creek River Kilometer 2.5    ......    65

31.    Time variations of simulated dissolved, particulate and
      total alachlor concentrations at Four Mile Creek River
      Kilometer 2.5 .    .    .    .    .    .    .    .     .     .    66

32.   Vertical distributions of simulated  sediment concentrations
      at Four Mile Creek River  Kilometer 2.5 at  12:00 p.m.,
      July 10, 1971	68

-------
                           LIST OF FIGURES (contd)

33.   Vertical distributions of simulated particulate alachlor
      adsorbed by sediments at Four Mile Creek River
      Kilometer 2.5 at 12:00 p.m., July 10, 1971   ....     69

34.   Vertical distributions of simulated dissolved, particulate
      and total alachlor concentrations at Four Mile Creek River
      Kilometer 2.5 at 12:00 p.m., July 10, 1971   ....     70

35.   Time variation of predicted flow rate at Wolf Creek River
      Kilometer 5, during the three-year simulation period   .     .     71

36.   Time variation of predicted total sediment  concentration
      at Wolf Creek River Kilometer 5, during the  three-year
      simulation period  .........     72

37.   Time variation of predicted particulate alachlor concentra-
      tion per  unit weight of sediment at Wolf Creek River
      Kilometer 5 during the three-year simulation period    .     .     73

38.   Time variation of predicted particulate alachlor concentra-
      tion per unit volume of water at Wolf Creek  River
      Kilometer 5  during the three-year simulation period   .     .     74

39.   Time variation of predicted dissolved alachlor concen-
      tration at Wolf Creek River Kilometer 5 during the
      three-year simulation period .......     75

40.   Time variation of predicted total alachlor  concentration
      at Wolf Creek River Kilometer 5 during the  three-year
      simulation period  .........     76

41.   Time variation of simulated flow Rate at Wolf Creek River
      Kilometer 5 during May 26, 1973 to July 15,  1973 ...     77

42.   Time variation of simulated total sediment  concentration
      at Wolf Creek River Kilometer 5 during May  26, 1973 to
      July 15, 1973	78

43.   Time variation of simulated particulate alachlor concen-
      tration at Wolf Creek River Kilometer 5, during
      May 26, 1973 to July 15, 1973	79

44.   Time variation of simulated dissolved, particulate and
      total alachlor concentration at Wolf Creek River
      Kilometer 5 during May 30, 1973 to June 9,  1973  ...     80

45.   Longitudinal distribution of simulated total sediment
      concentration at 6 a.m., June 4  1973	81

-------
                           LIST  OF  FIGURES  (contd)
46.  Longitudinal distributions of simulated dissolved, par-
     ticulate and total alachlor at 6 a.m. June 4, 1973     .     .      82
47.  Variations of simulated particulate alachlor in the top bed
     layer accumulated during the three-year simulation period    .      85
48.  MATC line for alachlor	90
49.  LC50 - MATC curve for Lasso®	90
50.  Curve A - SERATRA cutoff value	96
51.  Curve B - MATC line	96
52.  Curve C - Lethality curve  least likely to indicate
     mortality ...........      97
53.  Curve 0 - LC50 - MATC conservative curve     ....      97
54.  Curve E - LC50 - alternative curve	98
55.  Curve F - LC50 - MATC curve for durations exceeding
     24 hours	98
56.  Curve G - LC50 - MATC curve for durations exceeding
     48 hours	99
57.  Curve H - LC50 - MATC exceedance for durations greater
     than 96 hours  ..........      99
58.  Summary of toxaphene curves   .......      102
59.  Log-normal probability plot for dissolved pesticide
     concentration at Wolf Creek River Kilometer 5      ...      109
60,  Log-normal probability plot for dissolved pesticide
     concentration at Four Mile Creek River Kilometer  5      .     .      Ill
61.  Gamma probability plot for dissolved pesticide concentration
     at Four Mile Creek River Kilometer 5	112
A.I  Summary of global exceedance  results     	      127
   "  A phylogenetic chart of fish  used in this report   .     .     .      153
                                     xn

-------
                               LIST OF TABLES


 1.   Number of events defined by concentration C-j with durations
      of dj or greater for data  in Figure 2        ....      29

 2.   Number of time steps for events defined by  concentration C-j
      with durations of dj or greater for data  in Figure  2        .      29

 3.   Compilation of events in order of occurrence for example
      data     ...........      30

 4.   Global Exceedance Summary    .......      31

 5.   Runoff comparison for field sites number  1  and 2  .     .     .      49

 6.   Comparison of hydrology coefficients    .....      53

 7.   Comparison of sediment coefficients     .....      53

 8.   Sources of meterological data for the test  applications of
      the methodology to Four Mile Creek Watershed      ...      58

 9.   Approximate time periods for plowing, planting and
      cultivating soybeans at Four Mile Creek Watershed      .     .      58

10.   Test conditions for sediment and pesticide  transport
      modeling ...........      62

B.I   Pesticide effect on fish and crustaceans     ....    133

B.2   Scientific and common names of fish used  in this report     .    160

B.3   Pesticide classes  .........    151

C.I   Summary of adsorption interactions in natural aqueous
      environments.      	    177
                                     XT

-------
                              ACKNOWLEDGMENTS

     This study was sponsored by the U.S.  Environmental  Protection  Agency,
Environmental  Research Laboratory, Athens, GA.   The authors extend  their
deepest appreciation to Mr.  Robert B. Ambrose at EPA,  Athens and Dr.  James
W. Falco at EPA, Office of Research and Development for  their guidance and
patience.  The authors also  wish to thank  Dr. John G.  Eaton at the  U.S.
Environmental  Protection Agency, Environmental  Research  Laboratory  Duluth,
Minnesota for providing much of the toxicological. information used  in the
development of the risk assessment procedure.
                                      xiv

-------
                                  SECTION 1

                                INTRODUCTION
     Pesticides have long been recognized to cause a serious nonpoint
source pollution problem.  The use of chemicals to control pests on agri-
cultural lands has increased dramatically in the  last three decades and  is
still rising.  Between 1966 and 1976, pesticide production increased from
450,000 tons to 640,000 tons.  More than half of  these pesticides were
used by agriculture (EPA, 1972).

     Many of these pesticides are highly toxic to fish and other aquatic
biota.  Four major factors are responsible for the detrimental environ-
mental effects which may result from the use of pesticides:  1) some of
them are very persistent, so that even very low levels of these pesticides
in agricultural runoff may be of environmental concern; 2) they may be
both acutely and chronically toxic to fish; 3) pesticides are sometimes
widely and indiscriminantly used more as a preventative measure than as  a
cure in controlling pests; and 4) some pesticides are found to bioconcen-
trate in aquatic organisms resulting in a potential danger to man through
the food chain.

     In order to assess the impacts caused by both direct and indirect
activities of man, technically sound methodologies are required to predict
water quality degradation.  Currently, there are  no comprehensive method-
ologies which can be used to evaluate the aquatic impacts which may result
from the application of pesticides.  Consequently, a definite need exists
for the development of such a methodology.  To meet this need, this study
was conducted to achieve the following objectives:

 1)  Develop a methodology to predict the occurrence and duration of pes-
     ticide concentrations in surface waters receiving runoff from agri-
     cultural lands, and

 2)  Develop a preliminary risk assessment procedure to predict the poten-
     tial damage to aquatic biota.

     In addition to the development of this methodology, the feasilibity
of fitting overland and instream pesticide concentrations with statistical
distributions was examined.  This task was conducted to determine if a
simpler statistically based pesticide assessment methodology could be
developed to replace the more detailed computer model based methodology.

     In this report, Section 2 discusses overall  conclusions and recommen-
dations of this study, including the applicability and limitations of the

-------
 methodology,  study  results  on  the  feasibility of  the  simplified  statis-
 tical methodology,  and  suggested future  studies.   Section 3 describes  how
 pesticides  applied  to agricultural  lands behave on watersheds  and  in
 receiving streams.   This  section also  briefly discusses  physical and chem-
 ical  characteristics of pesticides.  The pesticide assessment  methodology
 is  described  in  detail  in Section  4.   The methodology was applied  to the
 Four  Mile Creek  Watershed and  to Four  Mile  and Wolf Creeks in  Iowa to
 demonstrate its  feasibility and  limitations.   Results of this  application
 are discussed  in Section  5.  An evaluation  of the methodology  in light of
 the methodology  application is presented in Section 6.   Section  7  des-
 cribes  study  results on the feasibility  of  the simplified statistical
 methodology.

      This report also contains the  following three appendices:   1) Appen-
 dix A,  describing the step  by  step  application procedure of the method-
 ology;  2) Appendix  B, presenting a  grouping of pesticides and  their physi-
 cal,  chemical, and  toxicological characteristics, and 3) Appendix  C, indi-
 cating  distribution coefficients, Kd,  of various  pesticides.

      Companion  reports to this  document are Mathematical Model  SERATRA  for
Sediment and Contaminant Transport in Rivers and  Its Application to  Pesticide
Transport in Four Mile and Wolf Creeks in Iowa, User's Manual for  the  Instream
Sediment-Contaminant Transport Model SERATRA,  User's  Manual for EXPLORE-I: A
River Basin Water Quality Model (Hydrodynamic Module Only), and Frequency
Analysis of Pesticide Concentrations for Risk Assessment  (FRANCO ModeT).

-------
                                  SECTION  2

                       CONCLUSIONS AND  RECOMMENDATIONS
CONCLUSIONS

     The Chemical Migration and Risk Assessment (CMRA) methodology was
developed for this project and applied to the Four Mile and Wolf Creek area
in Iowa to examine its applicability and limitations.  The following con-
clusions were obtained from this study:

 1)  Application of the CMRA Methodology to the Four Mile Creek watershed
     and Four Mile and Wolf Creeks in Iowa demonstrates that it is capable
     of predicting migration and fate of pesticides from agricultural lands
     and in receiving streams and of assessing pesticide impact on aquatic
     biota.  However, because of the lack of data, the methodology was not
     fully verified.

 2)  More specifically, since the methodology uses the state-of-the-art
     overland and instream models, the Agricultural Runoff Management (ARM)
     model (Donigien  and Crawford 1976), SERATRA (Onishi et al. 1979b) and
     EXPLORE-I (Baca  et al. 1973) to continuously simulate nonpoint source
     pollution prcoesses of both dissolved and particulate contaminants,
     it is useful for evaluating both short- and long-term migration and
     fate of pesticides.

 3)  Hence the CMRA Methodology can be used as:
     a)   a scientific decision making tool for regulation of toxic
          contaminants including pesticides and toxic substances.
     b)   a research  tool to estimate relative significance of various
          transport and degradation phenomena.
     c)   a tool to evaluate the effectiveness of various management
          practices to control toxic chemicals in the environment, and
     d)   a means of  identifying amounts of gaps and future needs in the
          risk assessment of contaminants.

 4)  The CMRA Methodology can be applied to large river basins by segment-
     ing the drainage area into smaller catchments whose sizes are limited
     by ARM's applicability.

 5)  The methodology  is applicable to nontidal rivers, streams and narrow
     impoundments. However, because of the limitations of EXPLORE-I,
     which was used to provide hydrodynamic data to the instream sediment

-------
     contaminant transport model, SERATRA, the methodology has;  limited
     applicability to very small streams.  This limitation may  be elimi-
     nated by replacing EXPLORE-I with another appropriate hydrodynamic
     model.

 6)  The risk assessment procedure, coupled with the statistical analysis
     of instream pesticide concentrations by a computer program FRANCO and
     pesticide toxicity data provides a good scientific basis for the
     pesticide risk assessment.  However, due to a lack of extensive know-
     ledge on pesticide toxicity, the risk assessment procedure includes
     only direct effects of dissolved pesticide concentrations  on aquatic
     biota.

 7)  Since the CMRA Methodology requires rather extensive field and  labo-
     ratory data, applicability and accuracy of the methodology could be
     impaired by the availability of required data.

 8)  The CMRA Methodology is general enough to be applicable to not  only
     pesticides, but to other hazardous substances and toxic chemicals, as
     well.

 9)  Both dissolved and particulate pesticide concentrations at the  stream-
     edge and instream were found to be best expressed by log-normal dis-
     tributions.  However, because of the complexity of migration and fate
     of pesticides on land surface and in stream, the preliminary feasi-
     bility study concludes that the development of the simp'lifed statis-
     tical methodology requires an excessive amount of computational time
     to cover a wide range of variables involved in order to derive  general
     statistical functional relationships among them.  Feasibility of the
     simplified statistical methodology requires further careful
     examination.

RECOMMENDATIONS

     Examination of the applicability and limitations of the CMRA Method-
ology leads to the following recommendations:

 1)  While the capabilities of the CMRA Methodology were partially tested
     using the Four Mile Creek Watershed data, more comprehensive data are
     required to fully evaluate all aspects of the methodology.

     More specific recommendations related to each component of the  CMRA
Methodology are presented below:

 2)  The following three considerations are recommended to extend the
     overland model, ARM:

     a)   Modification of ARM will be needed to include the effects  of
          tillage operations on runoff volumes and peaks.

-------
    b)   Coupling of the ARM and SERATRA models can be  improved  by modi-
         fying ARM to simulate the erosion of different sediment  size
         fractions and by partitioning of the pesticide to each  sediment
         size fraction.

    c)   Reduction of the data requirements for the overland  pesticide
         modeling component can be achieved by determining how model
         coefficients obtained during the calibration of ARM  on  a small
         catchment change when the model is applied to  larger size
         catchments.

3)  Additional improvement and extension of the instream transport
    modeling can be achieved by conducting the following:

    a)   SERATRA should be extended to include hydrodynamic modeling  (this
         will eliminate the use of a separate hydrodynamic code  and thus
         delete the process of transferring hydrodynamic information  from
         EXPLORE-I or other hydrodynamic codes to SERATRA).

    b)   Although EXPLORE-I was used for this methodology to  supply
         hydrodynamic information to SERATRA, othr hydrodynamic models
         can be used to simulate water movements in rivers.   If  other
         hydrodynamic models are selected, the compatibility  of  the
         hydrodynamic models and SERATRA needs careful  attention.

    c)   TODAM, a simplified one-dimensional version of SERATRA  is recom-
         mended to be used for cases where vertical distributions of
         pesticides are not needed.

    d)   It is recommended that this methodology be extended  to  estuaries
         and coastal areas where SERATRA is not suited.  This can be
         achieved by replacing current EXPLORE-I/SERATRA models  by other
         models suitable to these environments (Onishi  and Wise  1978).

    e)   More detailed and basic field and laboratory studies should  be
         conducted to investigate cohesive sediment transport, pesticide
         adsorption/desorption mechanisms and chemical  and biological
         degradation of pesticides.

4)  Additional studies must be performed to assess pesticide  toxicity and
    migration pathways.  Specifically:

    a)   The effects of low-level i-ntermittent pesticide concentrations
         on resistance building or cumulative adverse reactions of the
         organism should be examined.

    b)   It is recommended that actual field testing results  be corre-
         lated with laboratory bioassays.

-------
    c)   The relationship between toxic effects of dissolved pesticide
         uptake and the effects arising from ingestion of participate
         pesticides should be determined.

    d)   A more meaningful representation needs to be established regard-
         ing the likelihood of adverse effects from concentrations
         between the LC50 (a median lethal concentration) curve and the
         MATC (Maximum Acceptable Toxicant Concentration) range.  An
         indicator such as an LC20 would provide more definition to the
         area greater than the MATC and less than the LC50 curve.

    e)   Where a high potential for food chain biomagnification exists,
         it must be identified and be included in the interpretation of
         risk assessment results.

5)  MATC studies and more complete toxicity testing should continue.
    Future pesticide registration should require fish bioassay data on an
    extended time interval such as 120 or 168 hours for slower-acting,
    cumulative, or persistent chemicals.

6)  An accessible compendium of up-to-date time-LC50 response data should
    be available as support for future risk analyses.

-------
                                  SECTION 3

                        GENERAL REMARKS ON PESTICIDES
     Pesticides are used to kill or  inhibit  the  growth  of  certain  undesir-
able organisms inhabiting a particular  region.   They may be  classified
into various categories such as  insecticides,  herbicides,  fungicides,
nematicides, and rodenticides, based on a target life form.   Each  pesti-
cide has different physical, chemical,  and biological properties,  depend-
ing on its chemical structure.   Often,  compounds of similar  structure  will
behave in a similar manner, thus allowing for  the  organization  of  pesti-
cides into general chemical family groups as described  in  Appendix B.
Some of these groups are the organochlorines,  organophosphates,  carba-
mates, phenylamides, phenoxyalkanoates, and  triazines.  Agricultural usage
accounts for about one-half of the total pesticide consumption  annually.
Other major uses include pest control in parks,  golf courses  and other
recreation areas, along highways, utility lines  and rivers,  and  in home
lawns and gardens.  The focus of this report is  limited to pesticides
applied to agricultural lands.

     Pesticide sprays and dusts  are  seldom pure  formulations  of  the  active
chemical compound.  Usually a wetting agent, emulsifier, or  filler is
present.  The low water solubility of many pesticides requires  some  addi-
tive to facilitate application.  Toxicity of a pesticide may  change
because of the differences in chemical  properties  of the additive  and  the
bond of the pure compound with the filler.   In some cases  an  additive  may
prove more toxic to a species than the  active  ingredient itself.

     Most pesticides are aerially applied by spraying or dusting a target
area.  Some of the pesticide remains in the  atmosphere  or  is  deposited
outside the target area during the spraying operation due  to  pesticide
volatility, ga-seous drift, and deposition.  Deposition  of  a  pesticide  out-
side the target area by rain or particle settling may result  in  contamina-
tion of water sources and increased risk to nontarget species.

     After a pesticide has been deposited on the ground, it may  be lost  to
the atmosphere by wind erosion of soil  which has adsorbed  the pesticide
and by vaporization or codistination of certain pesticides from moist
surfaces.  The rate of vaporization at  the soil-air interface depends  on
the concentration and properties of the pesticide as well  as  soil  mois-
ture, soil  clay and organic content, soil temperature, temperature and
relative humidity of the air at the soil level,  and the wind  speed over
soil surfaces.

-------
     Pesticides applied to a field exist in either dissolved or particu-
late form associated with soil particles.  Heavy rainfall shortly after
pesticide application may result in runoff and soil erosion which physi-
cally removes the dissolved and particulate pesticide from the target
area.  Light intensity rainfall promotes the leaching of nonsorbed pesti-
cides into the soil profile and may dilute the remaining concentration
beyond its effective level.  It may be possible for the pesticide to enter
ground water if the soil texture is very loose and precipitation is suffi-
cient, particularly where the water table is near the ground.  However,
most pesticide enters into a waterway as a result of surface runoff.

     The migration of a pesticide is influenced by the amount and inten-
sity of precipitation or irrigation, temperature and other climatic condi-
tions, pesticide solubility, soil clay and organic content, soil particle
size distribution, adsorptive qualities of the pesticide, and watershed
characteristics.

     The Teachability of a pesticide is determined by the soil texture and
cation exchange capacity, amount of organic content, amount and intensity
of rainfall or irrigation, the mechanical placement of the pesticide, and
the adsorptive properties of the pesticide.  Insecticides may enter the
soil by direct application or by washing off plants.  Herbicides are often
incorporated into the top few inches of soil for proper performance.
Adsorption onto clays and organic materials in the soil may inactivate
pesticide movement and render the individual molecules unavailable for
solubilization or for plant uptake.

     Common ways to express the adsorption/desorption mechanism of pesti-
cide with soil are to use the Freundlich adsorption/desorption isotherm or
the distribution coefficient, Kd.  The Kd value is the ratio of the amount
of the adsorbed pesticide to the unadsorbed dissolved portion.  An
in-depth discussion of distribution coefficient along with estimated
values of Kd for various pesticides in water bodies may be found in
Appendix C.

     The length of time a pesticide remains effective depends on its per-
sistence.  Overall persistence of a pesticide in soil or water is depen-
dent on physical removal as well as biological, chemical, and photochemi-
cal degradation.  In many cases, a more toxic but less persistent chemical
has less environmental impact than a less toxic but more persistent one.
For this reason many of the very persistent chemical pesticides are now
banned.

     Degradation occurs when the complexity of the chemical complex is
reduced by splitting the molecule in some way.  Primary degradation is the
minimum change necessary to alter the chemical's structural identity.
Acceptable degradation is the minimum change necessary to remove some
undesirable property such as toxicity.  Ultimate degradation  is the com-
plete breakdown of the organic molecule to water, carbon dioxide, and
inorganic elements such as nitrogen and chloride.  Biodegradation and  in
some cases, burning, are probably  the only naturally occurring processes
capable of ultimate degradation.

                                      8

-------
     Usually the  initial degradation  step  will  detoxify a toxic compound.
Occasionally the  degradation products will  also be  toxic and  could be more
deadly than the original formulation  such  as  in the case of aldrin whose
degradation product  is dieldrin.  lexicological  synergism of  a  compound
with its degraded product  has  been  found with malathion and its breakdown
product, diethylfumate.  Further  research  may reveal  other examples where
an increase in toxic effects is related to  such  synergism.

     The pesticide degradation rate in soil depends on  soil temperature,
moisture, strength of binding  by  soil surfaces,  soil  type,  meteorological
conditions, cover crops, soil  cultivation,  soil  microorganisms, method of
application, and  pesticide formulation.  Oxidation, hydrolysis, and
microbial enzyme  action are the most  common methods of  degradation.  Photo-
degradation, the  chemical  breakdown caused  by light,  may occur  where
pesticides have been applied to a moist soil  surface  for an extended
period of time without rainfall.  A chemical  that breaks down rapidly in
moist soils could be very  persistent  in dry soils.   Sometimes persistence
is listed in terms of the  effective half-life of the  compound as a rela-
tive indicator of chemical persistence.  However, as  with other persis-
tence values, the half-life is an estimate  based on a particular set of
soil and climatic conditions.  Persistence may  be changed by  the amount of
pesticide applied and the  type of formulation additive.

     The rate of  biological degradation is dependent  on environmental fac-
tors and the availability  of microorganisms whose enzymatic systems can
break down the pesticide molecular  structure.   For  this to be initiated,
the compound must reach the microorganism  or  sometimes,  more  specifically,
the enzyme site,  and it must not be lethal to the organism.   The organism
must have the enzyme system present or be  able  to induce one  necessary for
chemical degradation.  Without the proper  enzymes,  the  water  soluable com-
pounds will probably be excreted, and the  insoluable  compounds  may be
stored.

     A pesticide  entering  a waterway  arrives  either in  the  dissolved form
or particulate form  (those adsorbed by sediment) during runoff.  What hap-
pens to it after  that point depends on stream conditions and  chemical char-
acteristics.  Some pesticides  are very soluble  and  will  stay  dissolved in
the water.  Others will remain in suspension  or  be  deposited  to the river
bottom after being adsorbed by sediment.

     Persistence of a pesticide in the surface  water  depends  on its rate
of degradation and its rate of removal from the  stream.   A compound in the
water may be broken down by hydrolysis, oxidation,  photolysis,  or
microbial degradation.  Persistence is usually  greater  in static water
than in active flowing streams and  is influenced by salinity, pH,  dis-
solved oxygen, and amount of organic material present.   Pesticide concen-
tration is reduced by flushing downstream, volatilization into  the atmo-
sphere, adsorption onto the bank soil and  bottom sediment,  and  plant and
animal  uptake.  Reduction  in pesticide concentration  by sorption and
uptake may be temporary.  The  scouring velocity  of  flowing  streams may
resuspend the sediment, and concentration  changes in  the dissolved pesti-
cide fraction may favor desorption from suspended or  bottom sediment.

-------
Residues in plants and animals may also be recycled back into the stream
or transferred to predatory organisms.  If the pesticide degrades quickly
to a nontoxic species, its metabolites should pose little environmental
impact.

     The existence of a pesticide in a stream may have a measurable effect
on local fish and other organisms.  Toxicity may be direct in the form of
lethal or sublethal responses.  Indirect effects may be exhibited through
biomagnification of pesticides through the food chain.

     Parameters affecting a pesticide's toxicity to a given species include
temperature, turbidity, pH, dissolved oxygen, its concentration, and chemi-
cal loading of the streams.  These may act synergistically or antagonis-
tically.  The chemical properties of the pesticide themselves are of utmost
importance as they dictate reaction trends with the various water quality
factors.  Most of these reactions increase or decrease the fraction of
dissolved pesticides to the adsorbed, inactivated fraction.  For example,
most pesticides are more soluble at higher temperatures.  If raising the
temperature promotes a greater concentration in the dissolved phase, and
thus a higher concentration in contact with fish, toxic effects would be
expected to increase.  Changes in pH may have a similar effect by changing
the relative pesticide-sediment and pesticide-water affinities.  The elec-
trolytic loading of a stream will tend to reduce the amount of pesticides
that can dissolve in river water.  Any reduction in the amount of dissolved
oxygen in the water may hamper the organism's efficiency for oxygen uptake
and decrease its ability to tolerate other undesirable effects.  Suspended
solids can cause abrasion to or clogging of the gills, possibly impairing
respiration.  The Dependence of the effect of each of these factors can
be somewhat generalized but is actually unique to each pesticide (Guenzi
1974).

     Biological aspects also influence pesticide toxicity.  The life stage
of an organism is a crucial component as pesticides may be harmless to
adult members but lethal to embryo and fry.  This lethality is due particu-
larly to the method of intake and the metabolic process affected.  Meta-
bolic differences between fish strains of the same species may exhibit
variability in toxicity as will comparisons of resistant and nonresistant
species.

     Partially because of the lack of toxicological knowledge of particu-
late pesticides on aquatic biota, the dissolved pesticide fraction is
usually regarded as the main source of toxicant exposure.  It enters the
organism by crossing membranes of aquatic biota.  The effect of the pesti-
cide will depend on its concentration, duration in the water, and  its par-
ticular toxic properties.  Food sources may contribute to ingested pesti-
cide exposure.

     Some organisms can adapt to reduce toxic effects.  Mobile species may
sense the presence of a chemical and seek to avoid it.  Others escape harm-
ful effects through genetic resistance.  These latter organisms, however,
can be particularly hazardous to predator species if the pesticide they


                                     10

-------
resist bioconcentrates to a high degree.  Predators may receive  a  lethal
dose from a high concentration in the prey.

     As discussed above, the pesticide released to environment undergo  com-
plex interactions and it is very difficult to assess its  impact  on  aquatic
biota.  Mathematical models can be used to integrate many of  these  complex
mechanisms controlling transport and fate of toxic chemicals  in  the  envi-
ronment into a single framework so that chemical migration, fate,  and
potential risks can be more accurately evaluated.  Hence, the Chemical
Migration and Risk Assessment (CMRA) methodology was developed to  provide
planners and decision makers in government and industry with  a sound basis
for evaluating the effects of chemical production, use, and disposal.
                                     11

-------
                                  SECTION  4

             CHEMICAL MIGRATION AND RISK ASSESSMENT METHODOLOGY
     In this section the Chemical Migration and Risk Assessment (CMRA)
Methodology for determination of migration and fate of pesticide and for
risk analysis will be discussed.  Subsection 4.1 presents an overview of
the pesticide risk assessment and its components.  In Subsection 4.2, each
component of the methodology will be discussed in depth.  Step by step
instructions to use the methodology are presented in Appendix A.

4.1  GENERAL DESCRIPTION OF THE CMRA METHODOLOGY

     The CMRA methodology consists of 1) overland pesticide modeling,
2) instream pesticide modeling, 3) statistical analysis of instream pesti-
cide modeling results, and 4) risk assessment procedure.

     Figure 1 provides an overview of the CMRA methodology developed under
this study.  As shown in this figure, meteorological data, information on
pesticide application rates and practices and watershed characteristics are
input to the overland pesticide transport model, ARM, to predict runoff,
edge-of-stream sediment and pesticide loadings.  Using these overland simu-
lation results, hydrodynamic (EXPLORE-I) and sediment-contaminant transport
(SERATRA) models predict pesticide distributions in the receiving streams.
The simulation results of the in-stream modeling are then used to develop
frequency distributions of occurrence and duration of dissolved pesticide
concentrations by FRANCO.  Finally, the statistical results with pesticide
toxicity data are used to assess the acute and chronic impact of pesti-
cides on selected aquatic biota.

Overland Pesticide Modeling

     The evaluation of pesticide transport and fate in a receiving water
body such as a river requires that the quanity of pesticides contributed
to the water body from agricultural lands be determined.  Following pes-
ticide application, pesticide losses from agricultural lands can occur
through surface runoff, soil erosion, volatilization, degradation (micro-
bial, chemical and photochemical), and uptake by plants and animals.  The
relative significance of these  loss mechanisms is highly dependent on
environmental conditions, agricultural management practices, and pesticide
properties.  Surface runoff and soil erosion have generally been recog-
nized as the dominant mechanisms by which pesticides are moved to a water
body.
                                      12

-------
            REQUIRED INPUT DATA
         ANALYSIS
• METEOROLOGICAL DATA
• PROPERTIES AND APPLICATION
RATES OF CHEMICALS
• WATERSHED CHARACTERISTICS


OVERLAND CONTAMINANT
TRANSPORT MODELING
(ARM MODEL)
         CHANNEL CHARACTERISTICS
         SEDIMENT CHARACTERISTICS
         UPSTREAM FLOW, SEDIMENT
         AND CONTAMINANT CONDITIONS
            TOXICOLOGICAL DATA
              •  LC50 VALUES
              •  MATC
   INSTREAM CONTAMINANT
  MODELING (SERATRA AND
     EXPLORE MODELS)
    STATISTICAL ANALYSIS
   OF PREDICTED INSTREAM
CONTAMINANT CONCENTRATION
 IN STREAMS (FRANCO MODEL)
                                                   RISK ANALYSIS
          Figure 1.  Chemical Migration  and  Risk  Assessment (CMRA)
                     Methodology

     Selecting the most  appropriate  pesticide  loading  technique for incor-
poration into the CMRA Methodology requires  an understanding of the types
of impacts agricultural  chemicals can  have on  aquatic  biota.  Pesticides
have been found to cause both chronic  (or  long-term) and  acute (or  short-
term) effects.  Organisms which are  repeatedly exposed to low levels of
pesticides coming off of agricultural  lands  after storm events can  experi-
ence chronic effects. Those which inhabit  areas along  stream channels where
contaminated sediments deposit can also  experience chronic effects.  These
effects may range from changes in growth and behavior  to  the impairment of
reproductive activities.  Chronic effects  are  particularly important for
highly persistent pesticides  like organochlor ides. Acute impacts occur
most frequently when intense precipitation events take place within sev-
eral weeks of pesticide  application.   Under  these conditions, runoff pes-
ticide concentrations may vary by an order of  magnitude or more during an
event (Donigian et al. 1977), and can  result in a high rate of mortality.
Because both chronic and acute impacts are important,  a mathematical model
is required to predict both average  and  peak pesticide concentrations.  To
accurately predict peak  pesticide concentrations, a continuous simulation
model is required because of the importance  of antecedent conditions in
determining the distribution of rainfall between  runoff and infiltration
during an event.  Ellis  et al. (1977)  recognized  the influence of weather
sequences in their study of pesticides in the  runoff from watersheds in the
Great Lakes Basin.  A continuous simulation  model is also required  because
                                     13

-------
the availability of a pesticide for transport by runoff is related not only
to the initial loading, but also to previous environmental conditions which
affect the degradation and washoff of the pesticide following application.

     Therefore, based on the need for detailed, continuous simulation of
both short- and long-term pesticide loadings, the Agricultural Runoff Man-
agement (ARM) Model (Donigian et al., 1976; 1978) was selected for incor-
poration into the methodology.  ARM was developed for the Environmental
Protection Agency by Hydrocomp, Inc.  The ARM model is composed of the five
major components which simulate:  1) the hydrologic response of the water-
shed, 2) soil erosion, 3) pesticide adsorption and removal, 4) pesticide
degradation, and 5) nutrient transformation and removal.  The hydrologic
component of ARM was derived from the Stanford Watershed Model; versions
of this model have been successfully tested on over 50 watersheds in the
United States (Donigian et al. 1977).  While the sediment and pesticide
components of ARM have only been tested on several watersheds, testing of
most other nonurban, nonpoint models is even more limited.

Instream Pesticide Modeling

     Pesticides reaching receiving water bodies from agricultural lands
migrate in both dissolved and particulate forms.  Most of the pesticides
adsorbed from solution onto sediment will be concentrated.  This process
may create a significant pathway to higher trophic levels through the food
chain.  Sediment contaminated by pesticides may be deposited on the river
bed, becoming a long-term source of pollution through desorption and resus-
pension.  In contrast, sorption by sediment can be an important mechanism
for reducing the area of influence of these pesticides by decreasing dis-
solved pesticide concentrations.  Other important mechanisms which reduce
pesticide concentrations in streams are volatilization, and chemical and
biological degradation.  Chemical degradation of pesticides occurs through
hydrolysis, oxidation, and photolysis.

     To obtain accurate temporal and spatial distributions of pesticides
in streams, a mathematical model must include all the important mechanisms
of both dissolved and particulate pesticide transport phenomena, such as
advection, and dispersion of pesticides, interaction of pesticides with
sediment, and chemical and biological degradation of pesticides.

     Presently there are only a few tested transport models capable of
simulating both dissolved and particulate contaminants with sediment-
contaminant interaction (ORNL 1978; Onishi 1979a).  Because one of the
most extensively tested models  is an unsteady, two-dimensional (longitudi-
nal and vertical) sediment-contaminant transport model, SERATRA (Onishi
et al. 1976; Onishi 1977a; Onishi 1977b; Onishi et al, 1979b), this model
was selected for the study.  The model consists of the following three sub-
models coupled to include sediment-contaminant  interaction:  1) a sediment
transport submodel, 2) a dissolved contaminant  transport  submodel, and
3) particulate contaminant transport submodel.  The original SERATRA was
modified to include the mechanisms of contaminant volatilization, and
                                      14

-------
chemical and biological degradation caused by  hydrolysis,  oxidation,  photo-
lysis and biological activities.  The model calculates  changes  in  the con-
ditions of both river bed sediments  and deposited  contaminants.

     Because both SERATRA and EXPLORE-I use a  constant  width  along each
computational cell, the hydrodynamic modeling  portion of the  one-dimen-
sional general water quality model, EXPLORE-I  (Baca et  al., 1973), was
selected to provide discharge and depth data to SERATRA.   However, other
dynamic models (e.g., Fread 1973) could be used.  The EXPLORE-I code  is
applicable to rivers, estuaries, and oceans to compute  depth  and velocity
distributions.  Although EXPLORE-I itself is one-dimensional,  it can  cal-
culate the longitudinal and lateral distributions by dividing  the  simula-
tion area with linked channels.  EXPLORE-I has difficulties handling  very
shallow streams with steep hydrographs.

     Computed results from SERATRA are statistically analyzed  to obtain
the probability of occurrence and duration of  given pesticide  concen-
trations.  The current knowledge on the toxicity of pesticides  has not
advanced enough to fully utilize the detailed  two-dimensional  (longitudi-
nal and vertical) distributions of both dissolved and particulate  pesti-
cide concentrations computed by SERATRA for the pesticide  risk  assessment.
Hence only cross-sectionally averaged dissolved pesticide  concentrations
were statistically analyzed for the risk assessment.  However,  additional
pesticide toxicological data in the future may enable SERATRA model outputs
to be more fully used.

Statistical Analysis

     A new statistical analysis procedure was  developed to provide mean-
ingful summaries of instream dissolved pesticide concentrations for the
methodology.  The computer program is called FRANCO (FRequency ANalysis of
Concentration program).

     The statistical characterization of the outputs from  SERATRA  was
designed to provide risk assessment information on  the  frequency and  dura-
tion of specified dissolved pesticide concentrations at selected instream
locations.  The outputs to be summarized are time series of dissolved pes-
ticide concentrations.  The procedure horizonally slices the concentration
versus time plot at a specified concentration  level, counts the number of
peaks (or excursions) above the line, counts the amount of time spent above
the line, and tabulates the peaks according to their duration.  This  pro-
cess is repeated for a range of concentration  values.  This is the compu-
tational basis used to determine the frequency and  duration of specific
pesticide concentrations.

     A concept of global exceedance of an LC-function (i.e., lethal con-
centration) was also developed for the risk assessment procedure.   Labora-
tory studies on fish provide basic information on the concentration of a
pesticide required to kill fifty percent of the population when exposure
durations are 24, 48, 96 hours, etc.  This information may be used  to
define a function between concentration and duration such  that  if  a point
is above the concentration-duration curve at least  fifty percent of the

                                     15

-------
population is killed.  These LC-functions are used to summarize both the
number of distinct events (i.e., peaks) and the amount of time spent above
the curve during the total simulation time.  This procedure was developed
to eliminate possible double counting of concentration episodes that would
result in lethality.

     The method used by FRANCO is one possible choice of bridging the gap
between simulation results and risk assessment procedures.  Similar pro-
cedures can be defined having the basic concepts used by FRANCO.

Risk Assessment

     This assessment is limited to the direct effects of dissolved pesti-
cides.  Ingestion as a second route for toxic effects is not addressed nor
are indirect effects such as bioconcentration and biomagnification.

     Toxicological assessment of damage to aquatic life is divided into
considerations of lethality and sublethal effects resulting from acute
and chronic exposures.  Acute lethality usually occurs within the first
96 hours (Sprague 1969), whereas effects after 96 hours are often con-
sidered chronic.  However, the actual dividing time between the two is
rather arbitrary.

     Important components of the risk assessment procedure include the use
of the median lethal concentration (LC50) data for predicting  lethality,
and the use of chronic data, maximum acceptable toxicant concentration
(MATC) values, and application factors for estimating sublethal toxicity.
These values can be used in conjunction with the mathematical models to
compute the extent of pesticide contamination in a stream and  the result-
ing probabilities of lethal or sublethal effects.

     The importance of lethality expressed as LC50, or the median toler-
ance limit (TLm) is paramount to the assessment process.  For  lack of more
adequate data, these values are applied to estimate the likelihood of fish
kills due to particular toxicant concentrations and durations  at lethal
levels.  The results are expressed as a probability.  Use of the LC50 is
justified by its common and historical usage as a parameter of toxicity.
The LC50 is a more reliable indicator than LC5 or LC100, for example,
because the variability of the average toxic range is usually  quite nar-
row, while the distribution of fish kills at the extremes tends to be much
more variable.  Use of the LC50 does, however, limit predictions of fish
kills to median lethality rather than threshold lethality for  any single
member of the population.

     Chronic toxicity is usually recorded as the amount of toxicant caus-
ing measurable effects.  These effects are typically described by symp-
toms.  Effects resulting in a decrease in growth, maturation,  or survival
of the individuals, an increase in physical deformities, a decrease in
reproductive ability demonstrated by a decrease in the number  of eggs, and
the reduced hatchability of the offspring are the criteria most often used
to assess chronic toxic effects.  Results from such bioassays  are either
descriptive, associating a given effect with a particular concentration,

                                     16

-------
or they are cited in terms of MATC range for the chemical.  The  lower  value
in the MATC range is the highest concentration tested which produced no
defined adverse effects.  The higher range value is the  lowest tested  con-
centration which resulted in a chronic effect.

     Sublethal effects may result from concentrations greater than the
MATC range if exposure time is sufficient.  Chronic bioassays are useful
as indicators of the minimum amount of toxin producing certain deleterious
effects on the organism.

     The amount of toxicant required to produce a sublethal effect differs
with the amount of exposure time.  For example, a species may show no
effects from a ten-day exposure to a chemical but may sustain damage when
exposed for 30 days at the same concentration level.  The species may  also
suffer reproductive impairment at some future point.

     The mathematical model requires the input of a sublethal indicator
for prediction purposes.  MATC values are based on  long-term tests which
may overestimate toxicity when applied to shorter time frames.   However,
unless the researcher has access to chronic data for all time steps under
evaluation, use of the MATC values as a sublethal boundary  indicator is
the best alternative available.

     While long-term toxicity testing and the use of MATC values are
becoming more common, many pesticides have not yet  been  tested for this
parameter.  Those for which MATC data is available  focus on a small number
of species because of the complexities of life cycle testing.  Where MATC
values are not available, they may be estimated by  the use  of an applica-
tion value which can be calculated or assigned arbitrarily.  The statisti-
cally significant mathematical relationship that appears to exist between
most LC50 values and MATC ranges is the basis for the use of application
factors which can be used to calculate MATC ranges  for other species.

     LC50 and MATC values are used to estimate various lethality and sub-
lethality curves.  These curves are used as input to the statistical
analysis code FRANCO.  The in-stream migration model (SERATRA) output  is
evaluated for concentrations exceeding the curves.  Global  exceedance  or
lack of concentration-duration periods exceeding the curves forms the  basis
for interpretation of the hazards from the particular pesticide  to the
organisms under evaluation.  However, risk assessment output from FRANCO
must be carefully reassessed by taking into account all  other known infor-
mation on the pesticide characteristics, life stage of the  species of  con-
cern, possible avoidance, reactions to the pesticide and other pertinent
data.

4.2  DETAILED DESCRIPTION OF CHEMICAL MIGRATION AND RISK ASSESSMENT
     METHODOLOGY

Overland Pesticide Modeling

     Various techniques are available for estimating the migration of
pesticides from agricultural lands.  A relatively simple technique based

                                     17

-------
on loading functions was developed by McElroy et al.  (1976).  This tech-
nique involves the use of the Universal Soil Loss Equation to estimate the
average soil loss from a catchment or watershed.  Since nutrients, organic
matter, pathogens and pesticides are commonly adsorbed to eroded  soil mate-
rials, pollutant loadings can be estimated by multiplying sediment yields
by factors which denote the concentrations of these substances  in the soil
and the effect of enrichment during the erosion process.  In general, the
approach is best suited to the prediction of annual average pollutant loads
expected to occur over many years.

     On the other end of the spectrum of pesticide loading techniques,
there are a series of mathematical models which simulate the complex pro-
cesses affecting pesticide cycling in the environment.  Examples  of these
pesticide transport and runoff models include the Agricultural  Runoff
Management Model, ARM (Oonigian and Crawford 1976; Crawford and Donigian
1973); the Agricultural Chemical Transport Model, ACTMO (Frere  et al.
1976); the Agricultural Watershed Runoff Model, AGRUN (Roesner  et al.
1976); and the Simulation of Contaminant Reactions and Movement Model,
SCRAM (Adams and Kurlsa 1976).  Mathematical models such as these are gen-
erally capable of simulating:  1) the complex hydrologic processes which
act to convert precipitation to runoff, 2) the erosion of soil  materials
due to rainfall impact and runoff, 3) the partitioning of pesticides
between dissolved and particulate (i.e., pesticides adsorbed to sediment)
phases, and 4) the loss of pesticide from the soil through volatilization
and degradation.

     Pesticide transport and runoff models are generally of two types:
single-event simulation models or continuous simulation models.  In single-
event models, the above processes are simulated only  throughout the dura-
tion of a runoff producing event.  Continuous simulation models,  on the
other hand, not only simulate these processes during  an event but also take
into account the processes which act to change soil moisture and  soil pes-
ticide conditions in-between storm events.  As a result, it is  not neces-
sary to estimate antecedent conditions prior to the modeling of runoff,
sediment and pesticide contributions.

     The pesticide loading model selected for incorporation into  the CMRA
Methodology is the Agricultural Runoff Management  (ARM) Model.  The basic
philosophy of ARM is that the processes controlling runoff, soil  erosion
and pesticide transport on the land surface are continuous  in nature.  The
status of antecedent soil moisture, land surface,  and pollutant conditions
are updated continuously even though nonpoint source  pollution  takes place
only during runoff-producing storm events.  As a result, the ARM  model
uses a continuous simulation approach.

     The basic hydrologic component of the ARM model  was derived  from  the
Stanford Watershed Model (SWM) (Crawford and Linsley  1966).  The  algo-
rithms used in SWM and the modifications which have been made to  it
throughout the development of ARM have been reported  by Oonigian  and
Crawford (Oonigian and Crawford 1976; Oonigian et  al  1977).   Basically,
the hydrologic component of ARM simulates the dominant physical processes


                                     18

-------
which act to convert precipitation  into runoff  (i.e.,  interception,  infil-
tration, interflow, percolation and evapotranspiration).  ARM  also has  the
capability to simulate the mechanisms of snow accumulation  on  the  land  sur-
face and runoff due to melting of the snow pack.  The  transfers  of water
between the major components of the hydrologic  cycle simulated by ARM are
not only used to calculate runoff,  but are also used in the calculation of
soil erosion and the vertical movement of pesticides in the soil profile.

     The soil erosion component of  ARM simulates sediment loss from  the
land surface due to both sheet and  rill erosion.  Soil fines are scoured
from the land surface and transported by overland flow.  The availability
of these materials for transport is a function  of rainfall  impact.   The
availability of soil fines is also  a function of the tillage operations or
conservation practices which have occurred on a watershed.

     Pesticides placed in a water-sediment mixture will eventually come to
equilibrium, with a certain fraction of the pesticide  adsorbed to the sedi-
ment materials and the remainder dissolved.  The adsorption and  removal
component of ARM uses a modified Freundlich adsorption/desorption  isotherm
algorithm to partition the available pesticide  between that dissolved in
the overland flow and that attached to soil fines being transported  by  the
flow.  Both single-valued and non-single-valued formulations of  this equi-
librium type algorithm have been included in ARM; these formulations are
discussed in detail by Donigian and Crawford (Donigian et al.  1977).

     The pesticide degradation component of ARM determines  the quantity of
pesticide available for transport,  either by overland  flow  or  by soil ero-
sion, at any time after its application to a watershed.  The major pro-
cesses which act to attenuate pesticides in the soil are volatilization and
microbial, chemical and photochemical degradation.  These processes, along
with plant uptake, remove most of the total pesticide  applied  to a soil,
with runoff and erosion removing only a small fraction.  In his  review  of
the literature on the pesticide content of overland flows,  Wauchope  found
that they generally contained less  than 0.5 percent of the  total applied
pesticide (Wauchope 1978).  Because reliable models have not been devel-
oped for each of these attenuation  processes, they are currently grouped
together in a simple first order decay algorithm in ARM.

     At the initiation of this study, the second version of the ARM
(ARM-II) was just being released.  As a result, the first version was
incorporated into the CMRA Methodology.  However, since the mathematical
models used in the methodology are  only input/output linked, ARM-II  can
also be used.  The only major differences between the models are that in
ARM-II:  1) the evapotranspiration  index parameters can vary on  a monthly
basis, 2) a soil compaction factor  has been added to represent the natural
aggregation and mutual attraction of soil particles and the compaction  of
the surface soil zone, 3) different first-order pesticide degradation rates
can be assigned to specific time periods after  pesticide application, and
4) chemical leaching factors have been added so that the amount  of pesti-
cide moving with infiltrated or percolated water can be changed.  These
differences are described in detail by Donigian et al. (1977)  and Donigian
and Davis (1978).

                                     19

-------
     The use of both versions of the ARM model is described  in detail by
Donigian and Davis (1978).  The major steps which are followed in the
application of ARM include:  1) data collection and analysis, 2) prepara-
tion of meteorologic data and model input sequence, 3) parameter evalua-
tion, 4) model calibration and verification, and 5) production of needed
information and analysis of simulation results.  The first three steps
basically involve the compilation and analysis of hydrologic and meteoro-
logic data and information on the watershed soil characteristics, topog-
raphy, cropping patterns and pesticide usage.  Specific data  requirements
and data sources are also reviewed by Donigian and Davis  (1978).  The
fourth step involves the calibration or adjustment of model  parameters to
improve agreement between simulated and recorded information.  This pro-
cess ideally requires that detailed data be available on  runoff, erosion
and pesticide contributions from the study area for at least three years
in order to accurately evaluate model coefficients.  The  final step
involves the application of ARM to the selected study area.

     One of the key limitations of the model is the amount of comparative
data required to calibrate model coefficients.  In order  to  test the CMRA
Methodology developed in this study, numerous agencies were  contacted in
an attempt to locate a comprehensive set of data.  It was found that rela-
tively few complete sets of data are available and that most of the avail-
able data had been used in the testing and development of ARM-I and ARM-II.
This limitation has been reduced somewhat by the regionalization of many
of the hydrologic parameters.  The hydrologic component has  been applied
to over 50 watersheds throughout the United States (Donigian et al. 1977).
The data limitation has also been reduced by the use of pesticide adsorp-
tion/desorption and degradation algorithms with parameters that can be
estimated through laboratory and field studies; many of these parameters
can be found in the open literature or obtained from pesticide manufac-
turers.  Additional testing and application of sediment and  pesticide
algorithms may also serve to reduce the model's data limitations.

     A second limitation is that the ARM model is not formulated to handle
the effects of channel processes on runoff and sediment transport.  As a
result, the model is applicable only to watersheds that are  small enough
that channel processes can be assumed negligible.  While  the limiting size
varies with climatic and topographic characteristics, it  was found to be
approximately 2 to 5 km^ (or about 1 to 2 mi^) (Donigian  and Crawford
1976).  Because of this limitation, an application of ARM to a large
watershed requires a number of individual model applications to smaller
subcatchments.

     The overland modeling component is the first step in the CMRA Method-
ology.  Based on known or selected tillage practices and  pesticide appli-
cation rates, ARM is applied to the study area to obtain  a continuous his-
tory of runoff and edge-of-stream sediment and pesticide  loadings.  This
information is then used as input to the second step of the  methodology,
in-stream pesticide modeling.

     The length of the study period is primarily based on the availability
of meteorological data, the projected computation time for the methodology

                                     20

-------
and the desired representativeness and stability of the  statistical
summary of instream pesticide concentrations.

Instream Pesticide Modeling

     A number of general water quality models  (Baca et al.  1973;  Norton
et al. 1970; Leendertse 1970) could be used to simulate  pesticide migra-
tion, but none of them includes the mechanisms of  adsorption  of pesticides
to sediment, desorption from sediment, and transport, deposition  and  resus-
pension of particulate pesticides associated with  sediment.   Computer sim-
ulation models which calculate pesticide transport without  including  the
pesticide-sediment interaction predict that pesticides will be flushed from
surface waters at the same rate at which water is  exchanged.   In  reality,
however, sediment sorption effects cause some  pesticides to flush much more
slowly, e.g., at the approximate rate at which the sediment is exchanged
in the surface water system  (EPA 1978).  Hence, in some  cases, neglecting
these sediment effects prohibits accurate prediction of  pesticide migration
(Onishi and Wise 1978).

     In order to include sediment-pesticide interactions for  the  instream
pesticide modeling, the sediment-contaminant transport model,  SERATRA, was
selected for this study.  The hydrodynamic modeling portion of EXPLORE-I
is used to provide discharge and depth data to SERATRA.  Input data for
EXPLORE-I are channel geometry and Manning's roughness coefficient,
together with initial and boundary flow conditions.  Outputs  of EXPLORE-I
are unsteady, cross-sectionally averaged distributions of velocity  (or
discharge) and depth.  Using the cross-sectionally averaged results pro-
vided by EXPLORE-I, SERATRA distributes a velocity vertically by  assuming
either a logarithmic velocity distribution or  a uniform  distribution.  A
description of the EXPLORE-I model formulation and user's manual  were
prepared by Baca et al. (1973) and Onishi (1979e).  Although  EXPLORE-I was
used for this study, other compatible hydrodynamic-water quality  models
such as QUAL-II (Norton et al. 1970) may also  be used to provide  hydro-
dynamic data to SERATRA.

     SERATRA is an unsteady, two-dimensional (longitudinal  and vertical),
sediment-contaminant transport model which utilizes the  finite element
computation method with the Galerkin weighted  residual technique  (Onishi
et al. 1976; Onishi et al. 1979b)).  The model uses the  general convection-
diffusion equations with appropriate boundary  conditions.   It consists of
three submodels coupled to include the effects of  sediment-contaminant
interaction.  The submodels are: 1) a sediment transport submodel,  2)  a
dissolved contaminant transport model, and 3)  a particulate contaminant
transport submodel.  A detailed descr-iption of the SERATRA model  formula-
tion and a user's manual were prepared by Onishi and Wise (1979c; 1979d).

     Sediment transport submodel:  The sediment transport submodel  solves
the migration (transport, deposition and scouring) of cohesive and  nonco-
hesive sediments.  The migration of three sediment size  fractions can be
modeled since the movements and adsorption capacities of sediments  vary
significantly with sediment sizes.  The submodel includes mechanisms  of:


                                     21

-------
 1.  Convection and dispersion of cohesive and non-cohesive sediment,
 2.  Fall velocity and cohesiveness,
 3.  Deposition on the river bed,
 4.  Resuspension from the river bed (simulating bed erosion and armor-
     ing), and
 5.  Sediment contributions from tributaries (contributions from overland
     runoff and other sources may be treated as a part of the tributary
     contributions).

This submodel also calculates river bed conditions, including changes  in
river bed elevation and the distribution of each sediment component within
the river bed.

     Dissolved and particulate contaminant transport submodel:  Both the
dissolved and particulate contaminant transport submodels include mecha-
nisms of:

 1.  Convection and dispersion of dissolved and particulate contaminants,
 2.  Adsorption (uptake) of dissolved contaminants by both suspended sedi-
     ment and bed sediment, or desorption from sediments into the water,
 3.  Chemical and biological decay and degradation resulting from hydroly-
     sis, oxidation, photolysis, and radionuclide decay where applicable,
 4.  volatilization
 5.  Deposition of particulate contaminants on the river bed or resuspen-
     sion from the river bed, and
 6.  Contaminant contributions from tributaries and other sources (contri-
     butions from overland runoff, wastewater discharges, fall out and
     groundwater are treated as a part of the tributary contributions).

Effects of water quality (e.g., pH, water temperature, and salinity) and
sediment characteristics (e.g., clay minerals, and organic content) can be
taken into account by changing distribution coefficients (which partition
the contaminants between particulate and dissolved forms) and contaminant
transfer rates to reach equilibrium conditions assigned by distribution
coefficients.  Transport of contaminants which are attached to sediments
is solved separately for those adsorbed by each sediment size fraction.
The contaminant distribution within the river bed is also computed.

     SERATRA is applicable to rivers and impoundments.  Since SERATRA  uses
a marching solution technique along a longitudinal direction, a reverse
flow condition (such as tidal flows) cannot be handled.  One of the advan-
tages of SERATRA is that it can be applied to water bodies over large  lon-
gitudinal distances and shallow depths.  SERATRA has been applied to a wide
range of hydraulic conditions, including the large, heavily regulated  Col-
umbia River (Washington) (Onishi et al. 1979b) and the intermediate size
Clinch River (Tennessee) (Onishi 1977a). Application of SERATRA to swift
streams such as Cattaraugus and Buttermilk Creeks (New York)  is currently
underway.
                                     22

-------
Statistical Analysis

     The  usefulness of  the  simulated  instream pesticide concentrations
depends in part on how  they are  statistically summarized for risk assess-
ment.  The statistical  analysis  bridges  the  gap  between SERATRA output and
the risk  assessment procedure.   The risk assessment  procedure estimates
the probability of lethality or  sublethality affecting  a given species
from aquatic exposure to a  particular  pesticide.   SERATRA provides a con-
tinuous prediction of cross-sectionally  averaged  dissolved pesticide con-
centrations at selected river segment  locations.   These two parts of the
methodology are linked  by the statistical  summarization computer program
FRANCO.   FRANCO (Frequency  ANalysis of Concentration) and its summary
methods were developed  especially for  this methodology.  A detailed
description of FRANCO and a user's manual  was given  by  Olsen and Wise
(1979).

     Statistically summarizing the simulated pesticide  concentrations
requires  a precise interpretation of  the phase "frequency of occurrence
and duration of pesticide concentrations."   The  risk assessment procedure
provides  information on pesticide concentration  that will result in an
effect if the concentration remains at or  above  a specified level for a
certain length of time.  This concept  forms  the  basis for the statistical
summary.  The summary of the  simulated concentrations uses specific defi-
nitions for these terms:  event  and duration of  an event.  An event defined
by a concentration of C occurs when a  sequence of simulated concentrations
begins below C at time  step t]_,  becomes  and  remains  greater than or equal
to C in subsequent time steps and then drops below C after time step t£
(see Figure 2 where C is Cs).  This is referred  to as an event defined by
concentration C.  The duration of an event defined by concentration C is
the length of time from the  beginning  to the end  of  the event, e.g.,  t2-t]_.

     Utilizing preceding concepts of an  event FRANCO provides three types
of summaries:  counts of time steps, counts  of events and LC-function
global exceedance.  The later is defined subsequently.   Let C\ C? •••  Cm
be a set of concentration levels used  to define events.   These are based
on lethal concentration information from the risk assessment procedure
and the concentration range expected from  the SERATRA simulation.  Let
di d£ ••• dn be a set of durations selected  on the same basis as the
concentrations.  All  of the  statistical  summaries are completed for events
defined by concentrations Ci(i = l,2,...m) with durations of dj (j =
l,2,...n,) or greater.  Each  of the m  concentration  levels  selected con-
ceptually slices the SERATRA  concentration time history plot horizontally.
A number of concentration peaks occur  above  the horizontal  line.   FRANCO
counts the peaks (events),  counts the  time above  the line and computes
summary statistics based on these counts.  Additional detail  is provided
by completing the above only  for events  lasting a duration  dj  or greater.

     The first summary  is a count of the number of events defined by
concentration C-j  with a duration of dj or  greater.   This is denoted as
NE(C-j,dj).  Events defined  by concentration  C-j of any duration,  i.e.,
greater than or equal  to one  time step,  are  denoted  by  NE(C > C-j).   This
is equivalent to NE(C-j,di).

                                     23

-------
                                                                               LTl
                                                                               CXI
                                                                               CD
                                                                               C\J
                                                                                               o>
                                                                                              -a
                                                                                               O
                                                                                               I/)
                                                                                               cu
                                                                                               Q.
                                                                          o


                                                                          

                                                                          o

                                                                          •I—



                                                                          CO  O
                                                                                                o  
                                                                                               i—  c:
                                                                                                Q- O)
                                                                                                E  o
                                                                                                ro  c
                                                                                                x  o
                                                                                               UJ  O
                                                                                                (U
                                                                                                s_
  o
  C\J
o
  LTi
  i-H

O
  CD
  i—i

O
  LTN    ra  cvj  r—i
O      O   O  <_>
                                 24

-------
     The second summary  is  a  count  of  the  number of time steps in all
events defined by concentration  Ci  with  a  duration  of dj or greater.  This
is denoted as NT(Cj, dj).   Also  NT(C > C-j)  is  equivalent to NT(C-j, d]_).

     The third summary measures  the frequency  of occurrence of concentra-
tions greater than or equal to C-j for  all  durations.   Mathematically,
this is given by the fraction:

                    NT(C  >C.)
      PT (C £C.) = 	T	—,  i = 1,  ..., m
where T is the total simulation  time  and  A is  the  time in a single time
step.  Pj (C >.C-j) is  a decreasing  function  of C-j  with PTJ(C >. 0) = 1 and
P-|-(C >. Cmax) = 0 where Cmax  is the  maximum simulated  concentration.

     As an illustration of these  summaries'  measures,  a particular time
series may indicate that for  10%  of the time,  the  concentration is greater
than or equal to C-j, i.e., Pj(C  >. C-j)  = 0.10.   At  the  same time the
total number of events that occurred  during  the fraction of the time may
be 5, i.e., NE(C 2 C-j) = 5.   For  another  case  a single event may be
responsible for the 10% figure,  i.e.,  Pj(C ^ C-j) = 6.10 and
NE(C >Ci) = 1.

     the fourth summary gives frequency of occurrence  information for
events categorized by  duration.   It is a  generalization of the previous
summary and is defined as:

         r              ,   ANT(Cisd  )
      PT  C > C., D >  d . I = 	!—*— .
       1  L     '       J J       T
This gives the fraction of the total  time  there  were  events defined by
concentration C-j that extended for  durations  of  dj  or greater.

     The following two relative frequency  summaries resemble conditional
probability statements and are useful when  additional duration  information
regarding events defined by concentration  C-j  is  of  interest. That is
why only one concentration level C-j  is  to  be  considered.   The summaries
are:
        r       ,        i   NE(Ci' di}
     PE [D >dj I  C >Ci] =NE(C  > g  )     and
                            NT(Ci'
                                     25

-------
They measure relative frequency with respect to  the  number  of  events
defined by concentration C-j and the amount of  time in  events defined  by
concentration C-j respectively.  Laboratory toxicity  experiments  provide
the main basis for developing a risk analysis  for fish.   A  common method
of summarizing the results of these experiments  is to  use a lethal  concen-
tration where 50% of the fish die.  A concentration, say  C96,  is deter-
mined such that 50% of the fish die when exposed continuously  for a speci-
fied duration of 96 hr, say d^6.  The previous summaries  give:

 a)  NE(C-j,d-j); the number of events having durations  and concentra-
     tions that exceed dj and C-j, respectively
 b)  Pj(D > dj, C >C-j); the fraction of the total time period  that
     the calculated concentration is greater than or equal  to  C-j for  a
     duration of dj or greater.

Usually information for LC50 concentrations at 24, 48  and 96 hours  is
available.  This information may be combined in  the  form  of an  LC50
curve.  In this study it is assumed that an LC curve is represented by up
to five peice-wise straight lines, as shown in Figure  3.  The  summary
NE(c96,  d96, C  >C96] gives the fraction
of the total time that the time-series  is in the 96-hr point exceedance.

     A more comprehensive measure of exceedance  is to  obtain the fraction
of the total simulation time that the calculated pesticide  concentration
exceeds an LC50 curve (shaded region above the LC50  curve in Figure 4).
This is termed global exceedance of the LC50 curve.  It is  defined  as:

                number of time steps in distinct events such that
                        (C. d.) is above the LC50 curve
    P (LC50) = <= - — J - —
     This summarization can be conducted  not  only for  an LC50 curve but
for other LC curves  such  as LC90,  LC10, etc.   An  example of the use of
this summarization involves deciding whether  fish will  be killed by acute
toxicity or chronic  toxicity  under  a certain  condition.   Assume that the
chronic toxicity has a duration of  96  hour  or more.   In  order to obtain
the answer, summarization will be  conducted for  the  following cases:
a) Case A has a LC50 curve shown by the solid line in  Figure 4, and b) as
shown  in Figure 5, Case B has a LC50 curve  which  is  the  same as the LC50
in Case A for the duration of 96 hour  or  more but is parallel to the ver-
tical  axis at d = 96.  PQ for Case  A provides the fraction of the total
time that 50% of fish will be killed by both  acute and  chronic: toxicity
of the pesticide.  Pg for Case B provides the fraction  of the total time
that 50% of fish will be  killed by only chronic  toxicity.  The difference
between the values of PQ  for  Cases  A and  B  will  then supply the frac-
tion of total time fish will  be killed by acute  toxicity.

     The remainder of this subsection  gives a detailed  example of the sta-
tistical summaries just described.  No new  information  is given.  However,
the step by step description  is useful in clarifying the concepts of point
and global exceedance.  The description is  given  in  the  order that the
                                     26

-------
                      24
                             96
                         DURATION
           Figure 3.
                 Example of LC50 curve with point
                 exceedance region
o
o
     .12
,24
,96
                12     24           96
                             DURATION
  Figure 4.  Example of LC50 curve with global exceedance region
                                27

-------
o

<

1
LLJ
O
Z
O
           ,24
            96
                       12     24
                                    96

                               DURATION
               Figure 5.  Postulated chronic toxicity global
                          exceedance region

computer program FRANCO completes the analysis.  The program documentation
and user's manual for FRANCO are available in Olsen and Wise (1979).
     As an example, suppose Ci, C2, 04, CIQ, Cis, and C2Q are the speci-
fied concentrations and di, d$, dio, dis, and d2Q are the specified  dura-
tions.  Table 1 presents the number of events defined by concentration
levels C-j and duration dj for the data given in Figure 2.  Table 2 pre-
sents the number of time steps occurring for all events defined by concen-
tration levels C-j and durations dj.  In the example di was selected  to
be one time step, GI to be zero, and C2 to be a model output cutoff  value
above which calculated instream dissolved pesticide concentrations are
statistically analyzed.  The cutoff value was introduced to FRANCO due
to accuracy limitation of the simulation model, SERATRA.  Under these
restrictions the total number of time  steps in the study period is given
by NT(Ci,di), the number of time steps with simulated concentrations above
the model cutoff by NT(C2,di) and the  number of events defined by concen-
trations above the model cutoff by NE(C2,di).

      In addition to the tabular summaries of numbers  of events and time
steps, a compilation of each  individual event defined by the  input concen-
trations C-j is kept internally  in FRANCO.  This compilation,  in the  order
that  it  is actually completed,  is presented in Table  3 to clarify  subse-
quent concepts.  Event concentration refers to the concentration  level  that
defines the event.

      The statistical summaries  available  from FRANCO  are discussed  as they
relate to issues of interest  in considering the  results  of  a  pesticide

                                     28

-------
  TABLE 1.  NUMBER OF EVENTS DEFINED BY CONCENTRATIONS C,  WITH  DURATIONS
            OF dj OR GREATER FOR DATA  IN FIGURE 2.
                                       Duration
                   '10
                   20
1
3
4
4
3
0
1
2
2
1
0
0
1
2
1
0
0
0
1
1
1
0
0
0
1
0
0
0
0
0
     TABLE  2.   NUMBER  OF TIME STEPS FOR EVENTS DEFINED BY CONCENTRATIONS
              Ci WITH DURATIONS OF
              FIGURE 2.
         dj OR GREATER FOR DATA  IN

dl
40
35
28
13
4
0

"5
40
33
24
6
0
0
Duration
dio
40
33
17
0
0
0
d!5
40
19
17
0
0
0
d20
40
0
0
0
0
0
                    15
                   C20
Simulation project.  A cutoff point  is used  in  the  SERATRA model  below
which the simulated concentrations are assumed  to be  zero.   The  number of
distinct events occurring with concentrations above the model  cutoff value
is given NE (03, dj).  There are three such  events  in the  example.   The
time over the study period during these events  was  35 time steps.   Note
that if the time steps are in terms  of hours, this  corresponds to  35 hours.
This information is also given as the percent of the  total  time  events
defined by G£ were present.
     More specific information is obtained by considering  the  duration of
the events.  For example, if a 10 hour LC50  value is  given by  Cq,  then
FRANCO gives the number of times this occurred, the total  time during these
occurrences, and the percent of time during  the study period this  LC50 con-
dition was exceeded.  The number of  events is given by NE  (Cq, d]_o),
the number of time steps in events defined by 04 and  d]_o °y NT (£4,
                                     29

-------
  TABLE  3.  COMPILATION  OF  EVENTS  IN  ORDER  OF  OCCURRENCE  FOR EXAMPLE DATA

     Ending Time Step      Number of Time Steps      Event Concentration

            5                        1                       C15
            7                        4                       CK,

           15                        2                       C1S
           17                        6                       C1Q

           19                       17                       C4
           20                       19                       C2

           23                        1                       C4

           24                        2                       C2

           26                        1                       C15

           26                        1                       C1Q
           28                        3                       C4

           34                        2                       C1Q

           37                        7                       C4

           39                       14                       C2

           40                       40                       Cl


and the percent of time by NT (C4, dio)/NT  (cl» dl)-  These
available output summaries from FRANCO are  valuable in the risk assessment
procedure as they give information on whether  the instream pesticide
concentration exceeds specific laboratory determined LC50 conditions.

     If more than one LC50 condition is of  interest, then each can  be  con-
sidered separately using the above approach.   In that case care in  the
interpretation of the results is necessary  because of possible double
counting of times and events.  An alternative  summary termed global
exceedance is given to eliminate double counting.  Global exceedance  is
based on the same concept as an LC50 function.  An LC50 function  can  be
defined by specifying pairs of concentration and duration values  based on
LC50 experiments.  By connecting these pairs with straight line segments
and extending the function  in a reasonable  manner at each end, a  function
is defined such that an event defined by a  particular concentration level
with a particular duration  can be classified as exceeding or not  exceeding
the function, i.e., exceeding an LC50 value (see Figure 6).  An event
exceeds the LC function when the concentration defining the  event and  the
duration of the event results in the pair falling above and  to the  right
of the function.  For the example data three events exceed LC  function 1
(see Figure 6):

     However, all three events occur during the same time period  of the
study, i.e., they consist of different slices  of the same pesticide con-
centration peak.  The global exceedance summary eliminates this double

                                     30

-------
                     TABLE  4.   GLOBAL  EXCEEDANCE  SUMMARY
     Ending Time Step     Number of Time Stepj
            19
            17
            15
                             17
                             6
                             2
                       Event Concentration

                               C4
                               C10
                               Cl5
o
I—
<
f—
UJ
O
o
o
            '20
            '15
            '10
                                            LC FUNCTION 1

                                   	LC FUNCTION 2

                                   	LC FUNCTION 3
                             I
'5           "10

         DURATION
                                                     J15
                                                           J20
     Figure 6.  Example of functions used for global  exceedance summary

counting by reporting only those events with the lowest concentrations that
occur in different pesticide peaks.   In this case only the event defined
by 04 is reported for LC function 1  in Figure 6.  A summary is given of
the number of non-overlapping events above the LC function, the time during
the study spent above the LC function, expressed as time and as a percent
of the total study period.

     By selecting different LC functions it is possible to differentiate
between short-term high concentration pesticide peaks and long-term low
concentration peaks.  By specifying  the LC function to be the model cutoff
concentration for all durations, as  in LC function 3, the location of each

                                     31

-------
pesticide occurrence instream can be determined as well as the global
exceedance summary information.  Figure 6 shows three LC functions which
split the concentration-duration space into four non-overlapping regions
(Numbered 1 to 4).  The percent of time spent in each region can be deter-
mined from the global exceedance percent summary.  First LC function 1  is
exceeded 42.5 percent of the time, LC function 2 is exceeded 45% of the
time and LC function 3 is exceeded 87.5 percent of the time.  Therefore,
during the study period the time spent in each region is:

                    Region 1             42.5%
                    Region 2              2.5%
                    Region 3             42.5%
                    Region 4             12.5%

     An interpretation of these numbers might be that 12.5% of the time
(Region 4) no significant pesticide concentration was present in the
stream.  For 42.5% of the time (Region 1), there are lethal impacts on
fish.  For 45% of the time (Regions 2 and 3), there are potential but
unknown lethal and sublethal impacts on fish.

Risk Assessment

     The risk assessment procedure provides a basis to predict the proba-
bility of lethality or sublethality affecting a given species from aquatic
exposure to a particular pesticide.  Two parameters have been chosen to
describe the risk analysis probability curves.  These are the use of LC50
data to generate the median acute toxicity curve and the maximum accept-
able toxicant concentration (MATC) to describe the effect-no effect bound-
ary for chronic toxicity.  Because of the difficulties involved with field
studies, such as keeping track of the organisms and eliminating food-chain
interactions, bioassays and toxicity testing have largely been confined to
special laboratory aquaria.

     Most laboratory procedures limit organism exposure to the water sol-
uble pesticide fraction.  Ingestion of pesticide on food particles is
eliminated as an exposure mechanism because the organism is usually not
fed during short-term tests.  The organism in its natural environment may
ingest pesticide sorbed onto food and debris surfaces.  Bottom dwelling
invertebrates and filter-feeding fish are also in contact with and may
ingest sediment and its corresponding adsorbed pesticide.  Ingestion as a
secondary route of intake is known in some cases to add to toxic effects
thus complicating direct comparison of field studies with laboratory
evaluations.

     An organism in the wild has many more potential environmental stresses
than the laboratory animal.  The field animal may be subject to starvation,
poor water quality, temperature changes, predation, and various chemicals
in addition to the pesticide of concern.  Its response to the pesticide may
be more or less severe than those in laboratory studies.  The additional
stress may act in combination with other factors enhancing the chemical's
toxicity; on the other hand, the organism may prove better able to toler-
ate or resist the poison because of prior adaptations.  Migration and

                                     32

-------
avoidance may allow mobile species to minimize exposure.  Consequently,
it is stressed that direct application of  laboratory results  to  field  situ-
ations must be made cautiously with an understanding of the underlying
assumptions involved.

     Investigators have used different techniques  in their measurements
of LC50 values.  The most accepted acute method  is the flow-through  test
where pesticide concentrations are actually measured.  Many static  bio-
assays have been conducted and some published test results are based on
calculations of pesticide concentration.   Results  from these  different
techniques are not always comparable.

     The Office of Water Planning and Standards, U.S. Environmental  Pro-
tection Agency has proposed the use of LC50 correction factors to stan-
dardize values to the 96-hour flow-through bioassay of measured  concentra-
tion (EPA, 1979).  While these factors are not final regulations and are
subject to change, this sort of approach may be  taken where standardiza-
tion of data is warranted.  The use of actual observed 24-, 48-, 72-,  and
96-hour LC50 data and of flow-through bioassays  of measured concentrations
is preferred when available.

     LC50 values derived from flow-through tests showed a geometric  mean
of concentrations to be 71% of those measured in static tests.   Conse-
quently, the Office of Water Planning and  Standards has suggested the  use
of this figure to standardize static results to  flow-through  conditions
(i.e., 0.71 multiplied by the static results approximates the concentra-
tion using the flow-through method).  Likewise,  their study of LC-50 values
based on actual measured chemical concentrations versus calculated  chemical
concentrations (based on the amount of toxicant  added to the  known  amount
of water rather than direct measurement after the  toxicant is in solution)
revealed an additional common discrepancy.  Calculated values tend  to  be
higher than actual measured concentrations.  Calculated data  may be  stan-
dardized to reflect the probable equivalent measured concentrations  by
multiplying calculated data by 0.77.

     In the same article the Office of Water Planning and Standards  offered
a procedure for standardizing 24-, 48-, and 72-hour bioassays to 96-hour
LC50.  This can be accomplished by multiplying these values by the  follow-
ing factors to estimate its 96-hour L.C50 data point.

     Given:  24-hr LC50, multiply by 0.66  to get 96-hr
             48-hr LC50, multiply by 0.81  to get 96-hr
             72-hr LC50, multiply by 0.92  to get 96-hr

Diverse pesticide and biological properties such as lag-time  and secondary
mechanisms of action are not considered.   Use of these factors implies a
fixed relationship between lethality and time.   While this assumption  is
not specifically correct for any particular pesticide, its use allows  for
relative comparison of toxicities cited in studies of different  duration.
                                     33

-------
     Construction of a median lethality curve requires  input of LC50  con-
centrations over a variety of time durations.  A time concentration mor-
tality (time-LC50) curve carried out to the  lethal threshold concentration
level is ideal for this use and should be applied when  available.  More
commonly, a lethality graph must be approximated from incomplete  sets of
LC50 data.  Toxicity data typically found in the literature are one or
more distinct LC50 values.  Finding values for specific  aquatic species
may prove to be difficult, or impossible, particularly  for chemicals  of
relatively low toxicity.  Appendix B includes tables of  pesticide toxicity
data for certain organic compounds.  Other data sources  include the EPA
Office of Pesticide Programs, EPA pesticide  registration files, and pesti-
cide manufacturers.

     The accuracy of the time LC50 curve improves by increasing the number
of LC50 data for different time intervals.   Confidence  limits  around  LC50
values are often listed and should be recorded for later use.  A  descrip-
tion of testing conditions should also be recorded with  the mortality val-
ues, as they may be useful in the selection  of the most  appropriate data
points.  When several investigators have reported conflicting  values  for  a
particular pesticide and species, the discrepancies may be the result of
specific conditions which may be discussed in the text  of the  article.
The LC50 value under conditions corresponding more closely to  case study
conditions would be the logical choice.

     Frequently, sufficient LC50 data do not exist to properly estimate  a
curve.  Figure 7 illustrates the procedure for estimating needed  FRANCO
input data points when these are not available.  When only one or two val-
ues are available for a particular species,  one of several methods can be
employed to approximate missing values.  The first of these  is to use the
slope of the toxicity curve of a related species for the same  chemical.
This principle is demonstrated on page A-5 in Appendix  A.  Appendix B pro-
vides a phylogenetic chart to indicate the genetic relationships  of various
fish species.  A species of the same genus is the preferred  choice for use
of slope data, though it may be necessary to use those  related at the
family level.  A second method for estimating missing data is  to  use  the
geometric mean of the toxicity slopes of other species  for the same pes-
ticide as shown on page 90.  Finally, the geometric means discussed above
in regard to LC50 standardization factors may be used to approximate
unknown values.  The reciprocals of the ratios correcting values  to a
96-hour LC50 can be used to estimate values  for 24-, 48-, and  72-hours
given a 96-hr LC50.  Note that these are approximations and  may  not reflect
any particular set of experimental values.   An example  of this method is
shown on page 91.

     In the event that  no data can be found  for a particular species,
an approximation of the curve may be derived from LC50  value trends given
for other species for the same pesticide.  Again, the selection  of  genus
related species  is most appropriate.  Assignment of a range  in terms  of
LC50 values is preferred where insufficient  data requires the  use of  other
family related species.  A geometric mean may be calculated  from  this range
for a more simplistic approach to probability modeling.  If  no values exist
for other species of the same family,  it may be useful, though perhaps less

                                     34

-------
            o
          = t;
                     So!
                     ££££
                     %i3o

                     -3:$-
                     £oS
                     < z ic
                     =;«3
                     asg
                     •f. <•
                     <_> Q
CULATE AND A
ED FROM REL
                                           isgs
                                           it g s LU
                                           a_  c£ ce:
                                           < tf o 
 ! => Q
 :gS
                       . QQ
                      o-  - £ u
                                          < o ~ —
                                          <£S5
                                          < ^y,£
                                          UJ T"
                                          a: S '
                                          
                                                                     O
                                                                     LT)
                                                                     CJ
                                                                     C
                                                                     en
£3 (^
C5 OO
S —
         ! •< i
                              i«;
Y VALUES WITH
PECIES CLASS
a; Q
o fc!
 -<
LO —I
i-
Sg
oo"<

52
5;:t

II
                                                                     -a
                                                                     a>
                                                                     u
                                                                     o
                                                                     CD
                                                                     S_
                                                                     3
                                                                     cn
OO C_J
LLl LU
Q- Q.
                         35

-------
justifiable, to use less taxonomically related species to determine  a
likely toxicity range.

     If no aquatic lethal toxicological data  is accessible for  a  particu-
lar pesticide, the use of data from other related pesticides may  be  an
indication of probable toxicity if biological data  shows similarity.  One
should seek out the same organism as is being studied for a toxicity range.
These approximation techniques only suggest trends  and are used solely for
model continuity.

     One further method describing the lethality curve is- presented, but
its proper application has not been determined.  It is based on the  appar-
ent hyperbolic character of the LC50 curve.  A computational method  of
determining duration-concentration values on this curve, particularly for
the lethal threshold concentration, is discussed by Dawson et al.  (1974),
and refers to a mathematical formula described by Wuhrmann (1952)  as

          (C-Cs)n(dm-ds) = K

where

      C = concentration of toxin
     Cs = concentration threshold
     dm = response time of fish
     ds = threshold of response time
    n,k = constants

The method may be limited to pure compounds or formulations demonstrating
only one mechanism for lethality.  Data points must lie  in an hyperbolic
pattern having both horizontal and vertical assymptotes  on a graph of time
versus concentration.  Use of this formula may have an important  applica-
tion as a method of extending the lethal curve beyond the last  experi-
mental point.

     The mechanism for predicting potential sublethal effects  is  based
on the use of the experimentally derived MATC range.  The MATC  range is
located between the highest value showing no detectable  harmful effects
and the lowest values displaying some observable effect.

     While recent focus on chronic data has greatly expanded its  avail-
ability, MATCs have not yet been generated for the  majority of  pesti-
cides.  The requirement for a MATC range is paramount to the sublethal
assessment process.  Figure 8 illustrates the alternatives when no MATC
has been published for a particular pesticide and aquatic species.  Where
considerable chronic data is available on long-term exposure, the MATC
range can be simply written as the interval between the  highest no effect
level to the lowest effect level.  If the tests were not run specifically
to establish such limits, this range may be wider than necessary.

     Where neither MATC nor adequate chronic  values are  obtainable for the
species of concern, data for other species' response to  the same  pesticide
                                     36

-------
   Identify Pesticide and
   Species of Interest
  Has a Specific
MATC been Renorted?
                                                     Yes
Enoloy this Value in Model
                                          No
                                  Is an MATC Available
                                  for Another Species
                                  for same Pesticide?
                                                     Yes
                             Use Application Factor
                            AF' = MATC//96 hr LC50
                             MATC = AF'x96 hr LC50
                           Prime indicates other species
                               Is the Pesticide Persistent
                                  or have Cumulative
                                    Toxic Effects?
                                                     Yes
                             Use Application Factor
                            AF = 0.01 for 24 hr Period
                            AF = 0.05 for any Point in
                               Time
                                          No
                                                              Use Application Factor
                                                             AF = O.D5 for 24 hr Period
                                                             AF = 0.10 for any Point in
                                                                Time
       Figure  8.   Procedure  for assigning chronic  toxicity input  data.

may be  applied.  The  existence  of  a correlation between  the  96-hour  LC50
and the  MATC has been noted  in  the Water Quality Criteria  -  1972 (WQC
1972).   An application factor  (AF) may be derived from known 96-hour LC50
and MATC values and may be applied to other  species exposed  to the same
pesticide to estimate its MATC.  The apparent  similarity of  the ratio of
MATC  to  a 96-hr LC50  for different species for a particular  pesticide has
been  discussed by  Eaton (1973).  The measured  MATC range is  divided  by the
96-hour  LC50 for the  known species to calculate an Application Factor (AF)
range.

           MATC Range/96-hr LC50  =  Application  Factor Range

The product of this AF and the  96-hour LC50  of the species of concern can
be used  to estimate  its MATC.

           AF Range x  96-hr LC50  =  MATC

      The following information,  as explained above, should  now be ready
for computer analysis of each pesticide and  aquatic species  under
consideration:

           Lethality:   24-hr LC50
                        48-hr LC50
                        96-hr LC50
                        Other LC50 points given  by toxicological studies
                        Incipient  LC50 if given  or calculated
                        Confidence intervals if  available.
                                        37

-------
       Sublethality:  MATC range
                      MATC point

     Ideally, all these points will be available for  the  graphing  of
lethality over both short and long exposures.  However,  in  addition to
the lack of one or more data points describing the LC50  concentrations  at
points 24, 48, and 96 hours, several assumptions are  required  to complete
the curve.  The curve must be defined between the point  on  the vertical
axis (C0, d0) and the shortest lethal data point and  also beyond the
longest lethal or sublethal data point.   Input of the minimum  time to
achieve a lethality response at high concentrations  (tm)  is  required.
Frequently this value lies between one and six hours, thus  allowing for
the assignment of a reasonable but arbitrary tm.  Sometimes  it may be two
days or more.  Experimental data should be used where available to resolve
this upper end of the LC50 curve.

     The LC50 line from d0 to the first data point may be most justifi-
ably dealt with by extending the LC50 curve from the  concentration at the
first known data point (Ci) horizontally  back to the  ordinate  (Figure 9).
This assumes that any value greater than  C0 (which equals (4)  exceeds the
LC50 line and is therefore included in the medium lethal  zone. This
assumption results in a conservative estimate for this section of  the
curve.

     A second approach for handling this  area assumes no  lethality exceed-
ing 50% occurs at any concentration for a species exposed less than time
equal to dj_.  It is represented by a vertical line extended  above  (C^.d^)
as illustrated in Figure 10.  This method is the least conservative mea-
sure of lethality for the duration d < d\.  The curve for this section
may be modeled in some alternate fashion  if enough evidence  exists to jus-
tify a different procedure.  An example of such manipulation is defining a
slope from some point C0 > Ci to point C\.  The lethal curve probably
approaches the ordinate at C  > C0 in a hyperbolic fashion.

     Straight line segments have been used to define  the  LC50  curve
between the known points.  An incipient LC50 value at a  given  time covered
within the time plot may be considered the final known point where it has
been given or calculated.  The section between the last  known  point and
the LC50 concentration at the arbitrary end of the curve  is  the second
area of uncertainty.

     This uncertainty has been dealt with in two ways.   The  first  proce-
dure is to extend the final known LC50 value horizontally to the end
(Figure 11).  This assumes the incipient  LC50 has been reached. Some-
times, however, the lethality curve continues a significant  negative  slope
past the last known value, especially when the effects are  cumulative.   In
this case the lateral extension of the last concentration value may vastly
underestimate the potential for lethality.

     The second procedure is a conservative lethality estimate. The  LC50
and MATC parameters are combined in a single relationship to assess the


                                     38

-------
          ppm
           14

           12

           10

            8

            0
      LC50
I	I	I
              0     24    48    72    96    120
                               Hours
Figure 9.   Horizontal  extrapolation of the LC50 curve to origin.
           ppm
            14

            10
      LC50
                     I	I
                  I	I
                     24     48     72     96     120
                                Hours
      Figure 10.   Vertical  extrapolation of the LC50 curve
                               39

-------
                ppm


                 14
                 12


                 10
LC50
                                      _L
              I
                          24
 48
 72    96

Hours
120   144
             Figure  11.   Horizontal  extension of the LC50 curve.

probability of short-term median lethality or long-term  effects.  This
curve is illustrated in Figure 12.  A value  of 96 hours  has been  selected
to merge the LC50 curve with the MATC line because  of its frequent usage
as a separation for acute and chronic toxicity.  Another time may be  pre-
ferentially selected for merging the curves  if the  LC50  curve is  known  to
level out at some point considerably before  or after 96  hours.

     FRANCO accepts a maximum of six pairs of concentration-duration
points to define each curve.  Where more than six pairs  are available and
necessary for modeling, these values may be  incorporated into two or  more
curves.  The areas exceeding the curves at all points or between  all  line
segments can then be estimated.

     The choice of the input curves depends  on the  needs of a methodology
user.  Some of the possible curves are illustrated  in Figures 13a to  13d.
Interpretation of the curves requires an understanding of the assumptions
they represent.  The curves a, b, c, and d   in Figure 13 are offered  as
the most informative at this time.  Additional curves may be generated  as
desired for the time frame deemed most important.   For example, the curve
may be described and its probability estimated for  a duration longer  than
48 or 96 hours.

     If no chronic data or MATC values can be obtained,  an arbitrary
application factor based on the chemical's persistence and acuTtiulation
tendencies as discussed in the WQC-72 standard may  be used.  WQC-72 defines
persistent as having a half-life greater than four  days.  If the  compound
is known to be persistent or be cumulative in its effects, use  0.01 as  the
standard application factor.  If it is neither persistent or cumulative in

                                     40

-------
                ppm

                 12


                 10
LC50
                                       I
              I
MATC

 I	
                          24    48    72    96    120  144

                                     Hours
       Figure 12.  Combination of the LC50 curve with the  MATC  value.

action, an AF of 0.05 should be applied.  Concentrations between  0.01  and
0.05 of the 96-hr LC50 should not be exceeded for more  than  24  hours for
the non-persistent chemicals.

     The MATC range for experimentally derived data may be converted to
a point for convenient computer manipulation.  Two options exist  for its
calculation.  The first is to take the geometric mean of both extreme
values as the MATC value.  The second is to choose one  of  the MATC  range
limits for the distinction.  Use of the highest no effect  value  is  the
most conservative method.

     The FRANCO output is a summary of concentration information  based on
certain physical and chemical characteristics.  It is an important  tool to
aid the interpretation of the potential effects of a chemical on  a  given
species.  However, there are other factors that may be  critical  to  a spe-
cific assessment.  The FRANCO output does not reflect biological  effects
other than toxicity and cannot reflect fish behavior in avoiding  or pre-
ferring the affected area, induced sensitivity or resistance to  the
chemical(s), life stages during which sensitivity is greatest or  least,
synergistic or antagonistic effects from other chemicals in  the water, and
a variety of other "real world" possibilities.  It is up to  those inter-
preting the results to consider these factors along with the concentration
summary provided by FRANCO to assess associated risks of pesticides.
                                     41

-------
    ppm
    1.5

    1.0

      .5

      0
              24    48    72    96   120   144   168
                                Hours
    Figure  13a.  The MATC line converted from MATC range.
1 1 1
1 1 1 1
     ppm
      12

      10
       6

       1

       0
         0     24    48    72     96    120    144    168
                                 Hours
Figure 13b.  Conservative estimate of  lethality  using MATC  line.
                               42

-------
         ppm




          14





          12






          10
           0
                         I	 I      1 	J	I	L
                  24    48    72    96    120   144    168



                                    Hours



         Figure 13c.  Least  conservative estimate  of  lethality.
         ppm




          14





          12





          10
          0
                  I	I
                 24    48    72    96   120   144   168



                                   Hours



Figure 13d.   Alternative  representation  of  lethality  in  unknown regions.
                                   43

-------
                                  SECTION  5

                         APPLICATION OF METHODOLOGY
GENERAL REMARKS
     The CMRA Methodology described in this report is the first attempt to
integrate mathematical models with statistical analysis and risk assessment
procedures to produce a comprehensive nonpoint source pollution assessment
tool for pesticides.  For this reason, it was necessary to evaluate the
soundness of the methodology by applying it to an agricultural watershed
and receiving streams.  In addition to evaluating the soundness, there were
other important reasons for the test application.  Mathematical-model-based
methodologies generally require large quantities of detailed field data.
Since data availability could be a constraint to the use of the methodol-
ogy, the test application was used to help identify data requirements.  A
test application of the methodology also provided a means of developing a
step-by-step procedure for its use (see Appendix A).  Finally, the appli-
cability, limitations, and required refinements of the methodology, as
well as each of its components, were best identified through such an
application.

     The first step in the application of the CMRA methodology was to
identify agricultural watersheds for which detailed field sampling programs
had been or are currently being conducted.  Representatives from the Envi-
ronmental Protection Agency, the United States Geological Survey, the
United States Department of Agriculture Science and Education Administra-
tion (previously the Agricultural Research Service) were contacted, along
with a number of state environmental agencies and university researchers to
identify potential data sources.  This survey showed that pesticide data
have classically been collected at two levels:  1) on large watersheds or
river basins usually several hundred square kilometers in size or 2) on
small field sites or test plots which are usually less than 5 to 10 ha  in
size.  Only recently have field data collection programs been initiated on
moderate size watersheds.

     One such program is being conducted by Iowa State University  (ISU) on
the Four Mile Creek watershed in central Iowa.  The basic objectives of the
field sampling program are 1) to collect data for use in testing and refin-
ing ARM and 2) to conduct complimentary model-oriented research related to
nonpoint pollutants from agricultural lands (Baker et al. 1979).  The
transport of sediment, pesticides and nutrients by water is being studied
in order to obtain a better understanding of the effect of agricultural
activities on water quality and to improve the usefulness of models as
tools for assessing alternative agricultural management systems.  This

                                     44

-------
three-year program, which began in 1975, involves cooperation among the
Agricultural Engineering and Agronomy Departments at ISU; Hydrocomp,  Inc.;
and the EPA Environmental Research Laboratory, Athens, Georgia.  Since the
data generated under the Iowa State University study generally met the data
requirements of at least the overland modeling portion of the methodology,
the Four Mile Creek watershed was selected for testing purposes.

     The Four Mile Creek watershed is located in northwestern Tama County,
Iowa and covers an area of about 50.5 km^.  It is typical of the heavily
cropped regions of Iowa, the major crops being corn, soybeans, oats and
hay.  Annual sediment yields for basins in this region are about 150  MT/km^
for basins on the order of 50 km^ in size and water yields average about
150 mm/yr (Baker et al. 1979).  Farmers usually plow corn lands and disk
bean lands in preparation for the next crop.  Fertilizers are applied by
all farmers and herbicides are applied by most.

     Figure 14 shows a map of the Four Mile Creek watershed along with
the locations of hydrologic and meterologic data collection stations.  The
sizes of the catchments being studied range from field sites on the order
of 6 ha in area to the entire watershed.  Precipitation  is measured by a
network of recording rain gages.  Other meteorological data are collected
at a weather station located near the watershed.  Runoff, either as stream-
flow or overland flow, is measured at each of the gaging stations.  In Four
Mile Creek, there are two gaging stations maintained by  U.S. Geological
Survey.  There is no permanent gaging station in Wolf Creek whose tributary
is Four Mile Creek.  Sediment, pesticides and nutrient sampling also  occurs
at each gaging station.  Details of the field sampling study, including
other aspects of the sampling program such as soil sampling for pesticides
and nutrients and crop inventories are presented by Baker et al. (1979).

     Since the purpose of the methodology application to Four Mile Creek
watershed was to evaluate the methodology and not to perform a complete
assessment, only a portion of the watershed was used.  This area is shown
in Figure 15 as hatched catchments along with the Four Mile and Wolf
Creeks; Wolf Creek is a tributary of the Cedar River.  For the purpose of
modeling runoff, sediment and pesticide contributions to Four Mile Creek,
the watershed area was subdivided into the three catchments:  the Gladbrook
(344 ha), the northern catchment (660 ha), and the southern catchment
(440 ha)(see Figures 14 and 15).  The movement of sediment and pesticide
were simulated over a 67.6 km reach of Four Mile and Wolf Creeks; the reach
extended from upstream of the Lincoln gaging station to  the mouth of  Wolf
Creek.

     The pesticide selected for the test application was alachlor or  Lasso
(trade name).  Alachlor is a preemergence herbicide used to control most
annual grasses and certain broadleaf weeds.  Alachlor is the most widely
used herbicide in the Four Mile Creek watershed (Baker et al. 1979).
Alachlor, applied at a recommended rate, has an average  effective soil
persistence of 6 to 10 weeks and is primarily degraded by microbial action
and chemical breakdown (Weed Science Society of America  1979; Stewart
et al. 1965).  While alachlor will adsorb to colloidal particles in the


                                     45

-------
1    o

B    s
00    =3
(L-5    <
z    °
3    s

                                                            ro

                                                            to
                                                            c
                                                            o

                                                           4->
                                                            n3
                                                           +J
                                                            CO

                                                            en
                                                            O
                                                                 •
                                                            00 "O
                                                            c  a;
                                                            o ^:
                                                            •i—  tO
                                                            +->  s-
                                                            «  cu
                                                            O 4->
                                                            o  f«
                                                            d)
                                                            s_
46

-------
                                               to

                                               o
                                               O)
                                               S-
                                              o
                                              -a
                                               c
                                               fO

                                               O)
                                               OJ
47

-------
soil, it tends to be transported from agricultural lands mainly in a dis-
solved form.  Its half-life in water depends on the aerobic character of
the water body and its sediments.  Alachlor on an aerobic sediment has a
half-life of probably less than two weeks.  Anaerobic sediment conditions
speed the degradation process and result in a half-life of about four days.

     Alachlor is moderate in its toxicity, causing acute effects at the 1
to 30 ppm range, and its level of bioconcentration is low for fish species
tested.  It has a solubility of 240 mg/1 and a Kd value of 10 to 50 ml/g.
As noted before, however, its persistence decreases with microbial inter-
actions with sediments.

     In order to obtain a long enough time history of pesticide concentra-
tions in Wolf Creek for statistical analysis and risk assessment, a three
year duration between June 1971 to May 1974 was selected.  More details
regarding the test application are given in the following discussions of
each component of the methodology.

OVERLAND PESTICIDE MODELING

     An evaluation of the ISU data showed that even though very few storms
occurred during 1976 and 1977 due to drought conditions, the data for ISU
Field Sites 1 and 2 (see Figure 14) were detailed and complete enough for
an initial calibration of ARM.  Since alachlor was applied only to ISU
Field Site 1 in 1976, this catchment was selected for use in model cal-
ibration.  Precipitation data from Rain Gauge No. 33 and meteorological
data from the ISU weather station were input to ARM along with other data
on catchment characteristics, crop growth and times of tillage and pes-
ticide application.  Initial values of model calibration coefficients
were established by using the guidelines given in the ARM User's Manual
(Oonigian and Davis 1978).

     Initial attempts to calibrate the hydrology coefficients on ISU Field
Site 1 showed that no single set of coefficients could be selected to rep-
resent the highly variable runoff conditions caused by soil tillage at the
site.  The dramatic impact of tillage practices can be illustrated by com-
paring the measured runoff from Field Site 1 with that measured at Field
Site 2.  Field Site 2 is located next to Field Site 1 (see Figure 14) and,
with the exception of cropping pattern, both sites have basically the same
size, topographic characteritics, soil composition, and precipitation pat-
terns.  Field Site 1 has an area of 5.7 ha and was planted with soybeans
in 1976, while Field Site 2 has an area of 7.6 ha and was planted with
corn.  In 1977 these crops were rotated.  Table 5 presents a comparison of
total runoff from both sites for each of the five major storm events which
occurred in 1976 and 1977.  Besides showing the impact that recent tillage
operations have on reducing surface runoff, the comparison also shows that
deeper tillage practices like plowing have a greater impact than disking.
Similar observations on the effects of various tillage systems on soil and
water loss from Iowa soils were reported by Laflen et al.  (1978).

     Future improvements to ARM should include the incorporation of till-
age parameters into the hydrologic component of the model.  Perhaps the

                                     48

-------
infiltration coefficient (INFIL) could be increased at the time of tillage
and then decreased to its nominal value through a linear or first-order
decay scheme.
         TABLE 5.  RUNOFF COMPARISON FOR FIELD SITES NUMBER 1 AND 2
 Storm
 Event

4/23/76
5/29/76
6/13/76
4/19/77
8/15/77
Rainfall
  (mm)

   28
   29
   37
   20
   38
                          Field Site #1
Runoff
 (mm)

  0.6
  3.8
  8.1
  4.5
  4.8
Tillage
Practice

Plowed
Planted
 Date

4/5/76
5/19/76
Disked    4/19/77

Runoff
(mm)
6.0
11.4
1.2
0.0
0.8
Field Site
Tillage
Practice
Disked
Planted
Cultivated
Plowed
#2
Date
4/2/76
4/28/76
6/7/76
3/31/77
     Figures 16 through 20 show comparisons between the measured  and  simu-
lated runoff for each of the five major storm events.  The measured runoff
for Field Site 2 and the average runoff for the two sites are  included for
comparison purposes.  These figures show that although the model  could not
be calibrated to exactly simulate each event, the model does simulate the
major trends fairly well.  Much of the difference between measured and
simulated values can probably be attributed to the impact of soil tillage.

     While this impact is highly evident on field size areas,  larger  catch-
ments (e.g., several hundred hectares in size) which are more  heterogenous
in terms of cropping and tillage patterns would not be as highly  affected
by tillage since all of the fields in a larger catchment would not be
tilled at the same time or to the same degree.  Since the ARM  model was
to be applied to areas larger than the Field Sites 1 and 2 in  the test
application, the calibrated hydrology coefficients for these field sites
were tested by applying ARM to the catchment area above ISU Station 8 (see
Figure 14).  This catchment is about 30 times larger than the  Field Sites
1 and 2 and closely approximates the sizes of the catchments selected for
the test application (see Figure 15).  While detailed runoff measurements
were only taken during 1977 at Site 8, comparisons between measured and
calculated flows indicated that several of the hydrology coefficients had
to be slightly modified.  The initial and modified values for  each coeffi-
cient are listed below.
                    UZSN
                    INFIL
                    INTER
                    Initial

                      0.15
                      0.10
                      0.70
                     Modified

                       0.40
                       0.03
                       1.50
     Leytham and Johanson reported an earlier application of ARM  to Four
Mile Creek watershed (Leyham and Johanson 1979).  Table 6 shows that their
calibrated coefficients agree quite well with the coefficients obtained
during this study.  The only exceptions are those hydrology parameters
                                     49

-------

-------
                     CO
                     E
1.0

0.8

0.6

0.4

0.2
                                                  FIELD SITE1
                                              —FIELD SITE 2
                                              - — AVERAGE
                                                  SIMULATED
                           10   20   30   40   50
                                      TIME (hrs-min)
                            8
                                 10   20
          Figure 17.  Simulated  and measured runoff  of the May 29,  1976
                      storm.
   0.12
   0.10
I  0.(
   0.06
   0.04
   0.02
                                      	FIELD SITE 1
                                      	FIELD SITE 2
                                      	AVERAGE
                                      •	• SIMULATED
30  40 50
10  20 30 40 50
10  20  30 40 50
 TIME (hrs-min)
                               10 20  30  40  50
                                                                          10  20  30
        Figure  18.   Simulated and measured runoff for  the June 13, 1976
                     storm.

-------
                                        FIELD SITE 1
                                  	FIELD SITE 2
                                  	 AVERAGE
                                        SIMULATED
                    10   20   30   40   50
                            TlME(hrs-min)
 Figure 19.   Simulated  and measured runoff for the April 19, 1977
             storm.
             0.4
             0.3
        o
        §5   0.2
        ex.
             0.1
                     	 FIELD SITE 1
                     	 FIELD SITE 2
                     	 AVERAGE
                           SIMULATED
   10   20   30   40    50
                                            10    20
                            TIME(hrs-min)
Figure 20.
Simulated and measured runoff for  the  August  15,  1977
storm.
                                52

-------

Coefficient
A
EPXM
UZSN
LZSN
K3
K24L
K24EL
INFIL
INTER
L
SS
NN
IRC
KK24
KV
Battelle-
Northwest
0.0
0.12
0.40
8.0
0.6
1.0
-
0.03
1.50
200.0
0.065
0.20
_
-
-
               TABLE 6.  COMPARISON OF HYDROLOGY COEFFICIENTS

                                             Leytham and
                                             Johanson(a)

                                            0.003
                                            0.09
                                            0.80
                                            7.8
                                            0.5(b)
                                            0.35
                                            0.0
                                            0.035
                                            3.5
                                          520.
                                            0.060
                                            0.23
                                            0.33
                                            0.97
                                            1.0


                (a) Average values for Segments 1, 2 and 3
                (b) Average values for April through October


which are related to the groundwater component of runoff (i.e.,  IRC, KK24,
K24EL, KV).  The ISU Field Sites were too small to calibrate these parame-
ters.  As a result, the coefficients that could not be evaluated were  sup-
plemented with Leytham and Johanson's coefficients.  Since no runoff data
were collected at either field site during the winter months, the snowmelt
coefficients could not be evaluated.  Therefore, Leytham and Johanson's
snowmelt coefficients were also used.

     Figures 21 through 25 also show the measured and simulated  sediment
mass removal rates for each storm event.  The model was calibrated to  the
Field Site 1 data since the two field sites were tilled at different times
and to varying degrees.  The sediment coefficients were calibrated to  be
consistent with the runoff results (i.e., if the simulated runoff was
greater than that measured runoff for Field Site 1, the simulated sediment
transport was also greater).  Table 7 shows a comparison between the sedi-
ment coefficients obtained under this study and those obtained by Leytham
and Johanson.  Again, good agreement was obtained.

     Soil-pesticide data collected during 1976 on Field Site 1 showed  that
the degradation rate for alachlor changed after the May 23, 1976 storm
event which occurred only ten days after application.  The degradation
rates before and after the runoff event were about 0.08/day and  0.06/day,
respectively.  The first version of ARM does not have the multiple degra-
dation rate option available in ARM-II.  A value of 0.06/day was selected
for the test application.
                                     53

-------
                               to
                               •>
                               o
                               in
                               in
                               c
                               eu


                               •5  o
                               o> +->
                               in  in
                               QJ r^
                               s_ en
                               3 t-H
                               O> ro
                               -a •—
                               C -I-
                               03  S-
                                   CL
                               -o <;
                               OJ
                               -»->  
                               •i—
                               u.
54

-------
                   c
                   E
     1600


     1400


     1200


     1000


b;   800


«   600


     400


     200
                  UJ
                  GO
                                                 FIELD SITE1

                                                 SIMULATED
                           10   20   30   40   50

                                       TIME(hrs-min)
                                  8
                                      10   20
         Figure 22.  Simulated and measured sediment mass  removal rate
                     for  the May 29, 1976 storm.
o
UJ
to
                                                                 FIELD SITU

                                                                 SIMULATED
    20 -
30 40 50
10  20 30 40 50    10 20  30  40  50

                    TIME (hrs -mini
                                                       10 20 30 40 50    10  20  30
         Figure 23.   Simulated and measured  sediment mass removal  rate
                     for the June 13, 1976 storm.
                                        55

-------
                                   FIELD SITE1

                                   SIMULATED
                17
                     10   20   30   40
                                 18
Figure 24.
               TIME(hrs-min)
Simulated and  measured  sediment mass removal rate
for the April  19,  1977  storm.
                                       FIELD SITE1

                                       SIMULATED
  Figure 25.
                10   20
                  TlME(hrs-min)
  Simulated and measured sediment  mass removal rate
  for the August 15, 1977 storm.

                     56

-------
                TABLE 7.  COMPARISON OF SEDIMENT COEFFICIENTS

                                 Battelle-      Leytham  and
                 Coefficient     Northwest      Johanson(a)
                   JRER              27?272

                   KRER              0.25           0.43

                   JSER              2.2            1.8

                   KSER              1.8            0.9

                   SRERTL            2.0            2.0
                 (a) Average of values for Segments  1,  2  and  3


     At the time of this study, data on pesticide concentrations  in  the
runoff from Field Site 1 were  limited.  In addition, no data  had  been  col-
lected by ISU on the adsorption and desorption of alachlor  on soils  found
in Four Mile Creek watershed.  Data such as these are  required  to deter-
mine the coefficients for the  Freundlich adsorption/desorption  isotherm
algorithm used in ARM.  Recent experimental work on  alachlor  adsorption
and desorption on a number of  different soils provided  values for the
Freundlich isotherm coefficients  (Jordan 1978).  For soils  of composition
and organic matter content similar to the silt-loams commonly found  in Four
Mile Creek watershed, K and 1/N in the Freundlich adsorption/desorption
isotherm alrogithm were found  to  be approximately 1.3  and 0.4,  respec-
tively.  Since only limited pesticide runoff data were  available,  these
coefficients were used without being tested.  Future comparisons  between
simulated and measured pesticide  runoff concentrations  are  required  to
verify the actual values of K  and 1/N for alachlor at  Four  Mile Creek
watershed.

     Following the calibration of the hydrologic, sediment  and  pesticide
coefficients, the ARM model was prepared for the test  application.   In
order to obtain a long enough  time history of pesticide concentrations for
statistical analysis, a three-year simulation of runoff,  sediment and
pesticide contributions to Four Mile Creek was performed.   Detailed  mete-
orologic data from the ISU study, however, were not  available for this
duration.  As a result, meteorological data collected  at  nearby weather
stations were obtained from the National Climatic Center  in Asheville,
North Carolina.  Table 8 lists each type of data required and the sta-
tion from which it was obtained.  The simulation time  period  selected  was
January 1970 through May 1974.  The first year of data  were used  to  calcu-
late initial soil moisture and sediment conditions.

     The dates for plowing, planting and cultivating were randomly selected
within the approximate time periods given by Leytham and  Johanson (1979)
for each field operation (see Table 9).  Soybeans were  assumed  to be the
major crop planted in the study area and alachlor was  applied to  the soy-
                                     57

-------
           TABLE 8.   SOURCES OF METEOROLOGICAL DATA
                     APPLICATION OF THE METHODOLOGY
                     CREEK WATERSHED
                FOR THE TEST
                TO FOUR MILE
                 Date Type
Weather Station    Period of Record
          Precipitation (hourly)    Traer, Iowa
          Pan Evaporation
Ames, Iowa
          Maximum and Minimum       Toledo, Iowa
          Air Temperature (daily)

          Wind Movement (daily)     Waterloo, Iowa
          Solar Radiation
          (daily)

          Dewpoint Temperature
          (daily)
Ames, Iowa


Waterloo, Iowa
August 1948 -
December 1976

January 1962 -
December 1973

April 1948 -
December 1976

January 1961 -
December 1976

July 1959 -
August 1972

January 1961 -
December 1976
              TABLE 9.  APPROXIMATE TIME PERIODS FOR PLOWING,
                        PLANTING AND CULTIVATING SOYBEANS AT
                        FOUR MILE CREEK WATERSHED
                  Plowing

                  Planting

                  Cultivating
    April 1 - May 1

    May 5 - June 1

    May 30 - July 1
beans the day after planting.  The only constraint on the selection of a
pesticide application date was that it could not occur on the same day as
a storm event occurred.

     In order to vary the rate of application of alachlor to the study
area, it was assumed that conventional tillage practices, similar to those
currently used on Four Mile Creek watershed, were used on the Gladbrook
and southern catchments and that conservation tillage practices were used
on the northern catchment.  The application rate of alachlor on the conven-
tionally tilled catchments was 1.8 kg/ha; this is the same application rate
used on ISU Field Site 1 in 1976.  Since the efficiency of weed control  is
usually reduced when conservation type practices were used, the alachlor
application rate was increased to 2.3 kg/ha for the northern catchment.

     Figures 26, 27, and 28 give a summary of the total simulated sediment
and pesticide contributions from all three catchments during the study
                                     58

-------
       700-

       630-

       560-
   u
   .£   490—

   g-   420-

   1   350-
       140-

        70-
        0-
                                                    TTT
                                                           T_r
    ii I  r i  | i  i ] T  r ]  i i  |  i' i  | i "i ') 11
 JUN  SEPT  DEC  MAR  JUN SEPT DEC  MAR  JUN SEPT DEC  MAR
 1971                1972                1973
Figure 26.   Time variation of  predicted sediment loading  to Four Mile
             Creek during the three  year simulation period
      5x10
  o
  8?
  o
  1
  oe.
  1
  o
  Lul

  O
         4.—
3.—
                                                   ilu_
            : TT | I  I  | I  I (  I I  |  I I  [ I  I  | I I  |  I I  | ' I  | I  l [  I  I [  I'
           JUN  SEPT  DEC  MAR JUN SEPT  DEC MAR JUN  SEPT  DEC  MAR
           1971                1972               1973
Figure 27.  Time variation of predicted dissolved alachlor loading to
            Four mile creek during the  three year simulation period
                                  59

-------
     <
     Q
     OH
     Q
     O
     O
        5X10"5-
            4.—
3.—
            2.—
            1 —
               : I  I |  I  I |  I I  | I  I |  I I  | I  i |  i i  |  I I |  I  I | I  i  | I  r
              JUN  SEPT  DEC MAR  JUN SEPT  DEC MAR JUN  SEPT DEC  MAR
              1971
                    1972
1973
  Figure 28.  Time variation of predicted particulate  alachlor  loading to
              to Four Mile Creek during the three year simulation  period

period.  Simulation results show that the major  runoff events at Four Mile
Creek watershed occur between May and September  each year  but not  in
winters or early spring.

INSTREAM PESTICIDE MODELING

     As shown in Figure 15, the study area for the  instream modeling  was a
67.6 km reach between River Kilometer 19.3 in Four  Mile Creek and  the mouth
of Wolf Creek.  Four Mile Creek joins Wolf Creek at River  Kilometer 48.3 of
Wolf Creek.  Runoff, sediment loading and pesticide loading from the  three
catchments calculated by the ARM model were introduced to  Four  Mile Creek
at River Kilometer 9.7.

     Unlike the overland pesticide modeling discussed  above, almost no mea-
sured data were available for the instream pesticide modeling.  Discharge
measurements of Four Mile Creek conducted at Traer  (River  Kilometer 4.2)
during the three year simulation period indicate the large variation  of
daily discharge ranging from 9.65 rtvVsec to 0.006 m^/sec.   The  average dis-
charge at Traer during this period was 0.46 m^/sec. Since the  purpose of
the methodology application to Four Mile and Wolf Creeks was to evaluate
the CMRA Methodology and not to perform an assessment  in this area, actual
flow variations were simplified for this study.

     The base flows of Four Mile and Wolf Creeks without runoff contribu-
tions was assumed to be 2 m^/sec and 7 m^/sec, respectively. Test conditions
                                      60

-------
for the instream modeling are shown in Table 10.  Based on field data
obtained by ISU (Baker et al. 1979), it was assumed that the river sedi-
ment consists of 65% sand, 15% silt and 20% clay.  For erMeo soil from the
overland, Baker et al. (1979) show a variation in a particle size distri-
bution as a function of time and sediment concentrati • for the 4/19/77
runoff event which occurred at Field Site 1.  Based on these particle size
distribution analyses performed by ISU on a selected runoff sampler, calcu-
lated total sediment loading from these three  j+chments were assumed to
consist of 0% sand, 40% silt and 60% clay.  Particle sizes of sand, silt
and clay were assumed to be 0.350 mm, 0.01  mm, and 0.0014 mm, respec-
tively.  No field or laboratory studies were performed to evaluate critical
shear stresses (^CRj and TCDj) and tne erodibility coefficient (Mj), needed
to calculate erosion and deposition of cohesive sediments (see detailed
model formulations of SERATRA ir Onishi and Wise 1979c).  These parameters
were determined through a trial and error calibration procedure.  The bed
shear stresses at each location for each time step were calculated by
SERATRA code internally with a known hydraulic condition.

     The distribution coefficient, Kd, of alachlor associated with the
bulk sediment was estimated to be 50 ml/g in receiving streams (Baker
et al. 1979).  In this study, Kd values of alachlor with sand, silt and
clay were rather arbitrarily assumed to be 2, 20, and 100 ml/g.  The total
pesticide alachlor for each storm event was then distributed to each sedi-
ment size fraction as follows:  A ratio of the selected Kd values of sand
to those of silt and clay is 1:10:50.  Hence for an equal amount of sand,
silt and clay, clay contains five times more alachlor than silt, which in
turn contains 10 times more alachlor than sand.  Since the sediment load-
ing from overland was assumed to be 40% silt and 60% clay, total particu-
late alachlor contributed from the three catchments was divided into 0, 12
and 88% each associated with eroded overland sand, silt and clay, respec-
tively at the stream edge.  Because of the lack of available data, the
transfer rate, Kj, for alachlor to moving sediment and stationary bed
sediment to reacn to an equilibrium condition assigned by a distribution
coefficient were arbitrarily assigned values of 0.36 and 0.01 per hour,
respectively.  Although SERATRA simulates various chemical and biological
degradation of dissolved pesticides individually, there were no available
data to enable the simulation of degradation by each process.  Hence in
this study, all pesticide degradation was lumped into a first order reac-
tion having a degradation rate of 0.003 per hour.  This rate was estimated
from preliminary information supplied by Monsanto Agricultural Product Co.

     Simulation of sediment and alachlor transport was conducted for each
of the following substances:  1) sand, 2) silt, 3) clay, 4) dissolved
alachlor, 5) particulate alachlor associated with sand, 6) particulate
alachlor attached to silt, and 7) particulate alachlor adsorbed by clay.
The modeling procedure for SERATRA involved simulating sediment transport
of sand, silt and clay.  The results were then used to simulate dissolved
and particulate alachlor by including the effects of alachlor-sediment
interaction.  Finally, changes in stream bed conditions were recorded,
including: 1) stream bed elevation change, 2) vertical and longitudinal
distributions of ratios of sand, silt and clay within the bed, and 3) ver-
tical and longitudinal distributions of alachlor within the stream bed.  A

                                     61

-------
TABLE  10.   TEST  CONDITIONS  FOR  SEDIMENT
              AND PESTICIDE TRANSPORT MODELING
      Base Flow, m°/sec
        Four Mile Creek                          2
        Wolf Creek                              7

      Manning Coefficient
        Four Mile Creek                          0.07
        Wolf Creek                              0.04

      Sediment Sizes, im
        Sand                               .     0.35
        Silt                                    0.016
        Clay                                    0.0014

      Bed Sediment Size  Fraction,  %
        Sand                                   65
        Silt                                   15
        Clay                                   20

      Size Fraction of Eroded Sediment from Overland,  %
        Sand                                    0
        Silt                                   40
        Clay                                   60

      Erodibility Coefficient, kg/m3
        Four Mile Creek
          Silt                                 1 x 10-*
          Clay                                 1 x 10-*

        Wolf Creek
          Silt                                 1 x 10-4
          Clay                                 1 x lO'4

      Critical  Shear Stress for Erosion,  kg/m2
        Four Mile Creek
          Silt                                  1.5
          Clay                                  1.7

        Wolf Creek
          Silt                                  0.55
          Clay                                  0.6

       Critical  Shear Stress for Deposition, kg/m2
         Four Mile Creek
          Silt                                  0.7
          Clay                                  0.6

         Wolf Creek
          Silt                                  0.4
          Clay                                  0.3

       Vertical  Diffusion Coefficient, m2/sec    1 x 1Q'6

       Distribution  Coefficient, m^/kg
         With  Sand                               0.002
              Silt                                0.020
              Clay                                0.10

       Transfer  Rate,  1/hr
         Moving  Sediment                          0.36
         Non-moving  Sediment                      0.001

       Total  Alachlor Degradation, 1/hr           0.003
                         62

-------
brief discussion of the simulation results will be presented  here.   More
detailed instream simulation results as well as mathematical  formulation
of SERATRA were reported by Onishi and Wise  (1979c).

     Calibrations of EXPLORE-I to obtain proper depth and  velocity  distri-
butions in Four Mile and Wolf Creeks were performed using  the predicted
largest runoff event (by ARM) which occurred during the three-year  simila-
tion period.  This event resulted in the maximum discharge of 18.1  m-Vsec
in Four Mile Creek.  The duration of the calibration run was  10  days  start-
ing from July 8 to July 17, 1971.  After numerous trial runs,  the proper
Manning coefficients in Four Mile and Wolf Creeks were determined to  be
0.07 and 0.04, respectively.  These Manning  coefficient values were judged
reasonable for these small streams.

     After ensuring mass conservation, EXPLORE-I and ARM simulation results
were input to SERATRA for the model calibration.  The only parameters  and
coefficients that require adjustment are the vertical dispersion coeffi-
cient and the three parameters,  i.e., erodibility coefficient, Mj and
critical shear stresses for erosion deposition, TCRJ and TQQJ used  to  cal-
culate deposition and erosion rates of cohesive sediments.  Transport  of
noncohesive sediments (sand) does not require the model calibration.   The
other parameters (e.g., pesticide distribution coefficients and  degradation
rate, sediment sizes, etc) are determined by theoretical and  experimental
analyses and field conditions prior to the model simulation.   Thus  most of
the calibration effort was oriented toward reproducing sediment  distribu-
tion patterns similar to the actual or estimated longitudinal  and vertical
distributions of sediment concentrations for the 67.6-km study reach.  For
model calibration, Four Mile and Wolf Creeks were divided  into fourteen
4.83-km segments.  A fifteen minute time step was used.  After trial  and
error, the values shown in Table 10 were selected for these sediment ero-
sion and deposition parameters.

     Time variations of computed sediment concentration for each sediment
size fraction and the sum of the three size fractions near the mouth of
Four Mile Creek (River Kilometer 2.5) are shown in Figure  29.  This figure
indicates that the majority of sediment is clay.  Figure 30 shows time
variations of computed particulate alachlor  associated with each sediment
size fraction, together with mean particulate alachlor concentrations
(weighted average'of three particulate alachlor concentrations associated
with sand, silt and clay) near the mouth.  Because clay has the  largest Kd
value, the particulate pesticide concentration associated  with clay has
the highest concentration.  Sand contains the lowest pesticide concentra-
tion.  Variations of computed dissolved, particulate and total (sum of
dissolved and particulate pesticide concentrations) alachlor  concentra-
tions with time near the Four Mile Creek mouth are shown in Figure  31.
The figure reveals that more than 75% of alachlor near the mouth of Four
Mile Creek was transported in the dissolved form and that  up  to  25% of
alachlor was carried by sediment.  This implies that sediment  transport
becomes important when a pesticide has a large distribution coefficient,
Kd, or a receiving water body contains very high sediment  concentrations.
                                     63

-------






in
OJ

£
i—
UJ
o
	 1

:*:
cc
UJ
;>
i — i
cc

^
UJ
cc
CJ

UJ
(—1
z:
cc
ZJ
o
i.











„ i
;
;
|
-
-

"




"~




_




—









—
;
-




t

1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i 1 i ! i i
1—
UJ
O
III
LU
CO
~J Q ^-
^5» -J Lam ^"
S ? =! 5
1— CO CO O
,
] 1

! i













/
o
-r— ^--*^

^^_— ^»* TT ^^^^ 	 	 ^fcr"
— g^~- 	 — 	 — ^ . ~S— 	 ";'
,|
[

1
|
Vt
V


1,1,1,1,1,1,1,1,1,1,1,1,1,1,

t
r—
i — i
: V0 ^f
i— i !^I
i/)
-a
c:
i— I to a)
o 5:

^ O OJ
2j -r- (U
'. 03 O
S-
-t-> (U
: c •—
O) •!-
c<^ o s:
• ^ r
cn u 3
fc o
CC -o LJ-
C3 <1J
j^j (^ (Q
'zs -M
.r- S
c/i £
•r—
•— i **- "°
| | O OJ
'. to
c f^ •
o 
; .r- 4-> •
r r~) +-> O C\J
" _^ fO "t"*
: "^ -r- s_
; S- -O O)
: > to aj
O) >, O
" 0s .^ ,2 ^1
; i— 'o 2
: °° a^
^_ OJ
: —i cu
— 3 3
' CD
in
     •   •   •    «
    *  
-------










in
•
CNJ
CC
LU
1-
£
O
_J
1 — 1
•xa
5
>
1 — 1
CC

*
LU
LU
CC
U
LU
__l
t—i
21
CC.
0
1 1
H-.









: ' 1
-
~-
'_
—
'- CZ.
'• O
: — i
- T
- M*M
:. o
: 3
: *£
• iii
: £
r 
>
CC
O
— i
x:
o
5

>
ct:
O
— i
x:
o
<:
-
<
— J
O
a:
f^ ^
:^
^
C£
O
__l
ZL.
0
-
«•!!•!
CNJtM(M(Mrj(n<»)(r)
OSSSQeSG
,111,11,
UJ UJ UJ It f tlJ LtJ li 1 QJ
oinotnscssQ
Qr^-in<\jsu7Qirj >
• •*<»>«*^
(vj--<— •--t'Htvjnro


-o
-2
.JH
o
o
(/>
ro
(/) O
C i—
O JC
•r-  
O 3 •
u o oj
•i —
S- 4-> S-
0 S- CD
S a. CD
o E
03 C O
i — rO i —
it! CD -r-
.— _, r0 -LJ Oj
u_ ___ ^_
° ^ 2 •?
O CcT
-i-5 CD .v
s- jc a>
rt> +J O)
CL. CD i-
Cno
•a o

03
C/l 4-1
O -r- C
•r- (/) O
-t-J •!—

S- C S-
fo fc -J-*
> l/l C
O)
a; ^ u
'"'" §


o
oo

ai
i.
=5
cn
G--01 X C9W/9W3 NOIiUUiN33N03
                              65
-------
                                                 CO
                                                 >-
                                                 <
                                                 o
                                                       LO
                                                     r"~  *
                                                       OJ
                                                        -p
                                                     •O CD
                                                      C E
                                                      
                                                      O  >
a ^
O- O)
   0)


si
10

•5 3
   O
-O U_
 0)
+J 4->
 (O 
-------
     Vertical distributions of sediment and alachlor concentrations  near
the mouth of Four Mile Creek are shown in Figures 32, 33 and 34.   In these
figures, elevations of water and bed surfaces are 272.56 m and  271.77 m,
respectively.  Figure 32 indicates that all suspended sediments have higher
concentrations near the river bed, but cohesive  sediments show  relatively
more uniform vertical distributions.  This is due to the smaller fall
velocities of silt and clay as compared to the fall velocity of sand.
Figure 33 presents vertical distributions of particulate alachlor  concen-
trations per unit weight of sand, silt and clay.  This figure reveals that
except those attached to sand, particulate alachlor concentrations are
almost vertically uniform.  Vertical distributions of average particulate
(weighted average of particulate alachlor associated with sand, silt and
clay), dissolved and total alachlor concentrations are shown in Figure  34.
In this figure, dissolved alachlor concentration is also shown  to  be almost
vertically uniform.  However, the particulate alachlor concentration per
unit volume of water indicates higher concentration near the river bed  due
to higher sediment concentrations near the bed.  The total alachlor  concen-
tration (sum of dissolved and particulate alachlor concentrations) reflects
patterns of dissolved and particulate alachlor distributions.

     Following model calibration, EXPIORE-I and  SERATRA were used  to simu-
late pesticide migration for the three-year duration between June  1971  and
May 1974.  For this case, the entire study reach was divided into  seven
equal-distance segments and a 30 minute time step was used to reduce the
required computational time.  Some of the simulation results near  the mouth
of Wolf Creek (River Kilometer 5) are shown in Figures 35 through  46.   Fig-
ure 35 shows the time variation of flow rate calculated near the mouth  of
the Wolf Creek.  The figure indicates that all 10 high flows occurred dur-
ing the three-year simulation period happened to occur in summer and fall.
Snow melt did not produce significant runoff to  receiving waters.

     Time variation of total sediment concentration near the Wolf  Creek
mouth is shown in Figure 36, which clearly indicates a small number  of
sharp peaks associated with high flows shown in  Figure 35.

     Figures 37 and 38 indicate time variations  of total particulate pesti-
cide concentrations associated with sediments per unit weight of sediment
and per unit volume of water, respectively.  There are two large peaks  of
particulate alachlor concentrations per unit weight of sediment (Figure 37)
but only one large peak of particulate pesticide concentrations per  unit
volume of water (Figure 38).  This is due to the fact that although
alachlor concentrations attached to sediment per unit weight of sediment
during the summer of 1972 is high, the sediment  concentration during the
same period is relatively low, so that the particulate alachlor concentra-
tion per unit volume of water becomes low.

     Dissolved and total (sum of dissolved and particulate) alachlor con-
centrations near the mouth of Wolf Creek are shown in Figures 39 and 40,
respectively.  There are two high peaks of dissolved and total  alachlor
during the three-year simulation period.  Figures 35 through 40 clearly
                                     67
-------
                                         CM
                                               g
                                    o

                                    o
CM
f~
CM
            CM

            CM
              o
               •
              CM
 CM           CM

uu 'NOI1VAJ13
                                                       o
                                                        fO
                                                        S-
                                                        CD  •
                                                        O O.

                                                        §0
                                                        o o
                                             0)
                                                       -a
                                                        CD LO
                                                        t/>   •
                                                          CM
                                                       -a
                                                        OJ  S-
                                                       +j  cu
                                                        (O -)J
                                                       i—  QJ
                                                           ^
                                                        <4-
                                                        O S-
                                                           

                                                        O     ^
                                                        i/i o) r^
                                                        •r- i— CT>
                                                        -O ••- •—

                                                        r~~      *^
                                                         cc s- o
                                                         US'—"
                                                        •^ o
                                                        4-> U_  >5

                                                         QJ •*-»  3
                                                         CM
                                                          a>
               'NOI1VA313
                              68
-------


-
-

-

1 1 1 1 1 1 1 1
«* CM O OO
3o
z P
o -5
~S>
^ o
o x;
i— —
°1S
CM CM CM i— H

CM CM CM CM
UJ 'NOIJ.VA313 ^
i-H









ill ill"
f-H CX
0
— ! en
ZC p
o -£
«g_ en
ift
0 ti! !H
^< ^^
— ^ C/1
-_J _•—
H f-
1 — ~»
o; >
*— > -**•
"* CM O OO *~" Q^
CM CM CM •— I
i*- i — i — i»~
CM CM CM CM
w 'NOI1VA313
i
^^
	 	 -




i i i i i i
So
^ ^ en
' < ^
UJ ~
«t  S-
(O QJ
f— >
o a: <-H
4-3 sx CT>
i- QJ «-)
rO CD
Q. S_ «
0 0
T3 i-l
CD O)
-(-> r- >,
fl3 'r— r^
1 ^3
•^ $~
co O •
u_ E
<+-
O 4-> D.
co O
c: co O
0 -!-> ••
•i- C CM
4-> QJ «— 1
••— -a res
i- CD
4-> (/> LO
CO •
•r- >>
o _a a)
•r- S_ E
+-> o o
S- CO i—
•
CO
CO
CD
^3
CD
U_




"J  NOI1VAJI3
            69
-------
                                                X Q£
                                                   3
                                                   3;
                                                   o
                                               CM Q
                                                            -O
                                                             c
                                                             CO
                                                             3 s-
                                                             o cu
                                                            •f- >
                                                            +J T-
                                                             s_ a:
                                                             (O
                                                         1
                                                   g
                                                   to
                                                   00
CM
 CM

 Pd            ^
 CM            «M

ui 'NOI1VA313
                                           oo
                                             •
                                           I—I
                                           f-
                                           CM
                                                                     >•>
                                                (O  -4-> r-
                                               i—  n3  3


                                                I  ss".
                                                i/> O   •'
                                                   •i-  E
                                                U-  4->   •
                                                O «3 G-

                                                (^ -(-> O
                                                ceo
                                                o ai  ••
                                                •r- O CVI
                                                +J C T-t
                                                3 O
                                                J3 O +->
                                                •r-     (O
                                                1_  S_
                                                +J  O ID
                                                VI i—   •
                                                •i- _G CM
                                                -o  o
                                                    03  S-
                                                .— i—  QJ
                                                 rtJ 
                                                 S_ 4-> i
                                                 (1) O •
                                                  0)

                                                  3
                                                  cn
                              70
-------

s
X
«-*
4M
co
£
i
»"4
t«k
«-«
X
CO
•
m
e
&
O
>M>4
*:

oc
i— «
or
*
£
B

u_
•"•"•
CJ
3*








••••^^H^HBM*
^MM«MMB«










•MBMMNHM
.a^TB!







































1 1 1 1 1 1 1 1 1
^D
O
-m ^;
L^J £: *-
U. °^ (U
•-1 >
H "r^
a:
, -^ -a
' — nj o
•• f •» Vf W
^ 0) T-
0 s_ s.
o cu
a.
M_
<— C
3- 5 °
§ :*"~
5 •JJ
^> j-> ro
(O r—
3
a> e
+J •(-
m M g ^
rs ^1
4_ i

O.
-»• cn
 S-
C 3
L o-o
4-> IT>
00 K •- «-
r",UJ 0^ i- <"
^ r-< « +J
> QJ

— ou o
•»— •! —
.t -^
0
*
to
- CO
O)
Z i-H 3
^j r~~ en
•"i en *i~
i— 1 U-
8 s s s 8 s
• • ^ • • • • Gp (9 Q Q O
cswo)
           71
-------
en

ID
CO


in

cc
LU

LU

o
cc
LU
 LU
 LU
 CC
                                                                               ID
                                                                               O


                                                                                o
                                                                                 CO
                                                                                •O
                                                                                 o
                                                                                          ro
                                                                                          C  I
                                                                                          O 3
fO l/l
s-

C ro
O) OJ
O >>
c:  i
o oj
o QJ
   S-
4-> JT
•i- JC

QJ
to CD
   c

rC ^»

o -a

   ur>
•a
QJ s-
+-s CD
0 4->
•i- QJ
-o e
QJ O
S- r—
O--r-
 o s-
   QJ

 §•^-6
•r- C£ O
+->    -r-
 fQ -i<: S~
•i- Q) QJ
 s- a> o.
 cO &-
 > CJ C
      o
 QJ <+- -r-

•^ 'o ro
                                                                                          UD
                                                                                          oo
                                                                                          3
                                                                                          CD
                         p-      CD     in      =j«      en     CM


                       C1/3WD  N01iyyiN30NOO  1H3WIQ3S
                                             72
-------






_4(
J^.
•x
CD
IT)
,!*
\
v^
CO

«
in
cc
UJ
UJ
o
_J
2
cc
UJ
>
cc
-
j~-
UJ
UJ
cc
'°
	 1
o







in











•














^
-•— M«=a^=s=^













•-D
S 1*- °
Li_ CTl -i-
I-H -i-> in
i- S-
c +->
CD O)
1- ° I
"° 35
~ o s-
r— O)
z: -c >
- => £ £
r, ^ Qi
rO ^£
QJ •
~ O) oi -o
+-> S- O
Q-, OO rt3 <_)•!—
UJ ' — ' — l-
- tZ cri 3 M- QJ
^H O •— O-
•r- 0
4-> 3 C
S_ 0
fO -fJ "f— '
CL fO -»->

"° SI i
0 E -r-
_ -1— •!— I/)
O) O) S-
— , i- (/) rO
_ ^ D. OJ
r° i«- o i
O O)
-(-> d)
- c -c s-
O O^ c~
s - S**3
~ it! cr, .<« 2 0)
r— H *' -p^
its •<-
— > c en
3 C
O) ••-
o .^ ^ s~
~ o i— CL-CJ
CO
^ r^ 0)
1 1 i 1 i 1 I 1 1 -00^^
                                                              CD
O)
                CD      =3<      00      <-i
              73
-------
en
in
CO
cc
ua
  _

o
 cc
 LLJ
 cc
 LU
 LU
 CC
 li-

 o

                                                                                                              §

                                                                                                    O
                                                                                                    O
                                                                                                    O
                                                                                                    o
                                                                                                              (O
                                                                                                              S-
                                                                                                                 un
         o cu
         C •(->
         O OJ
         u E


         o^

         JC
         O J-
         n cu
         r— >

            CC   •
         cu    -a
         i^ y/  ^|

                                                                                                              4-> 4-  C
                                                                                                              S- t—  O
                                                                                                              (O  O  ••-
                 <—
          tu  to g
          i %     C
             s_ .,_
                                                                                                               0
                                                                                                               aj

                                                                                                               D.
                S-
                (O
                O)
         >+- O  I
          o    a>
             CD <1)

          O =1 .C
en   CM
U-l    >
                                                                                                               (O -r-
                                                                                                               > c  cn
                                                                                                                  3  C
                                                                                                               a>    -r-
                                                                                                               E s-  v-
                                                                                                               •r- CU  3
                                                                                                               »— cx-o
                                                                                                               CO
               i         i
             O)       QO
                                         CO
                                                                     m
                                                                              CM
                                                       74
-------





sj*
p»^
^
(TO
in
\
CO

in
oc
UJ
LU
O
t— i
"*£•
CC
UJ

^-^
cc

t^
UJ
LU
CC
0
u_
	 1
o
3
















.










3
•M^^MS==













^
CO
LU
Lu



1—
-CJ
0

z.



CQ
LU
- Lu
_
| 	
- 0
O


§
f~3



CO
- LU
LU



!_
~ O


Z.
i 1 1 1 1 1 1 1 1 ^
C
_.. o
~ -r-
^ "rO
r* H »^
c s-
(i) it)
O OJ
c >>
O 1
O QJ
OJ
$_ J-
O ^
O O)
'fC
-o c
a; •.-
CV, > ^
^ S"°
1-1 CO LO
•o s_
OJ
T3 ••->
O)  T3
0 -r- 0
DC --
C S-
o -^ a>
•(- a) Q-
rO S- C
CM I°°
^ S£ »
-H Or-
O) 3 3
•r- +J -r-
i — a t/i
.
n
1
r-H CD
cn u

75
-------
                                                                                        -pp
00
\
ID
CD
CC
UJ
I—
UJ

o
cc
UJ
 UJ
 UJ
 cc
 o
 o
                                                                                           o
                                                                                           o
                                                                                                   J-> S-
                                                                                                   fO re
     O  O)
     C  S-
     o -E
     o +->
                                                                                                   o
                                                                                                      QJ
                                                                                                   o  en
                                                                                                   re  C
                                                                                          m   £   '-
0}  +J LO
    •O -!->
        •
    4-  >  T3
     O T-  O
       o:  -r-
     C     S-
     O J^  OJ
    •<—  
 cr>  >  i—  fC
                                                                                                    E    E
                                                                                                    •r- +J -i-
                                                                                                    i— ta 1/1
                                                                                         - O
                                                                                                5;
                                                    3QIOIlS3d
                                                 76
-------
in
CD

in

in

£
UJ

3
-<
                    1       1
                                                                                        .0
                                                                                         ro
                                      ID
                                      'CM
                                                                                        _o
                                                                                                    O>
                                                 O)
                                                 0)   •
                                                 S- 00
                                                O 1^
                                                   CTl
3 LD
   t—I

"to  >-,

cu  ^
-t-> 1~3
(O
S-  O


O 00

<4- (Tl
   f—I
T3


IT3 CM


E  co


    o^

O -r-


O T3


(O

C  O)
n3 +->
>  O)

O)  O
                                                                                                   2!
                                                                                         CM   OO  ,~'
                                CSHOD
                                                77
-------
ro

in
ro
CO
OJ
in
in

cc
UJ
i—
UJ

o
cc
UJ
UJ
cc
u
 CO
  Q
                                                                              in
                                                                             -in
                                                                             -in
                                                                              CO
                                                                             -in
                                                                              CM
                                                                              C\J
                                                                             -in
                                                                                       o
                                                                                      •i-  o
                                                                                       ro
                                                                                       s- ro
                                                                                       
                                                                                       o 10
                                                                                       O C\J
                                                                                      •i-  en
                                                                                      TJ  c:
                                                                                       (O
                                                                                       -(-> LO
                                                                                       o
                                                                                 £    oio)

                                                                                       rO  O
                                                                                       
                                                                                       O -i-
                                                                                         Qi

                                                                                       O ^ CO
                                                                                      •r- o) r-.
                                                                                      4-> <1> C^l
                                                                                       (O S- i—I
                                                                                      •i- CJ
                                                                                       S-      "
                                                                                       rO 4- LD
                                                                                      •r- -|_>
                                                                                      I— ro
                                                                              Q
         Q
                 S


                 CD
                        S


                        C^-"
S

in
Q

 •
=i<
Q


CO
©


OJ
Q


Q
                                                                                       |

                                                                                       en
                      C1/9M]  NOIlUyiN3GNOO 1N3WIQ3G  1B101
                                                 78
-------
en
in
00
CD
CM
\
in
in

cc
IU

LU
z
o
cc
UJ

I—I
cc
UJ
CC
o
                                                                          —o
>-
_J
ID
•"O
         d
         in
                                                                           ID
                                                                           "CNJ
    o
   -C\J
  -LO
                                                                          _o
                                                                           LT>
       00
             c:
             QJ  O
             O 4->
             C
             o oo
             o r-~
               cr>
             s- ^-i
             o
                                                                                     o 
         •o ^->
         cu  a)
         ••->  e
         ro  o
             V) s_
               CU
            14- >
             O •!-
               01
             C.
             0^:00
            •i- Q) I—
            +-> O) CTl
             fO S- i—I
            •i- O
             S-     «
             (O 4- IT)
             > i— r—I
               O
             cy 3 >,
                                                                                       to 1
                                                                                    OO
                                                                           C\J
                CNJ
                       in
                              O5
                                            CC
                                                   in
                                                          m
                                                                 in
§
               caw/aw]  NouyyiN33Noo  3aioiis3d  3iyinoiiyyd
                                             79
-------
                                                       oo
                                                              CO

                                                              <
                                                              Q
         CNJ
           •

         O
   CD
   C

 "3 S-
+-> 3
 O TD
4J
   in
-o
 C S-
 rO O)

 CO CO

 (O O

 3 T-
 O N/*
•r—
+-> S-
 5- CO
 rtJ >
 CX-i-
   o:
 •H
T3 ji<:
 CO CO
 > CO
<— S-
 o o
T3 O

•a
CO -l->
-(-> (0
     co
     r-
     en
                                                                    3  c en
                                                                    E  o
                                                                    •1- •!- CO
                                                                    CO .)-> C
                                                                       (O 3
                                                                    it-  s. "-o
                                                                    o +->
                                                                       c o
                                                                    c  a) •(->
                                                                    o  o
                                                                    •i-  C CO
                                                                    -t->  o r-~
                                                                    co  o en
                                                                    •i—    t—i
                                                                    S-  S-
                                                                    rO  O  "
                                                                    > i— O
                                                                      .C CO
                                                                    CO  O
                                                                    £  to >^
                                                                    •r- I— ITS
H/9W) NOI1VN1N30NOO 3aiOIIS3d
                         80
-------
^o
£
a
a:
o
CE:
O

*


i
L
L
C
c
L
«





—

1 —

,
a
£
J
L.
_l
»




1 1 1

C

-^ -i— ro
^r -05-
p 3 -(->
i£2 -r- 0)
o Q £?£
^ 00
_1 O
LO
^•N
*-^ 
-------
                                                                                                          O)

                                                                                                          (O
                                                                                                          S-

                                                                                                          Q.
C£
                                                                                                LU

                                                                                                o
         T3
          O)
          O
          CO
          CO
                                                                                                            OO
                                                                                                         0)
                                                                                                         £  c

                                                                                                         CO r-3
01
o
                                                                                       CM
                                                                                   JJ  C5
o
00
CO

O
 CO   •
 C  «3
 o
•i- VD

 3 4->
JD  tO

 5-  S_
4->  O
 CO i—
•i- .E
T3  U
 CO  (O
 C

•o  
.)->  O

 CD
 C  -O
 O  C
    i      i      i
                                                                                                         IO
                                               82
-------
indicate a small number of sharp peaks of water discharge, sediment  con-
centration, and concentrations of particulate, dissolved and total alachlor
occurred during the three-year period.  Comparison of Figures  38,  39 and
40 reveals that most of alachlor transported during May 1972 and May 1973
was in a dissolved form.

     There simulation results revealed that high pesticide concentrations
in the streams do not directly correlate with peak runoff or soil  erosion
events, but rather with the time between pesticide application  and the
first storm event after the application.  In other words, pesticide  con-
centrations in the stream have strong seasonal patterns (high  peaks  in
May and June) corresponding to pesticide applications that were closely
followed by precipitation causing runoff and soil erosion.

     The results imply that the amount of pesticide being transport  is con-
trolled by the supply of pesticide on the land surface.  Since  alachlor
degrades very rapidly after application to farm land, significant  improve-
ments in water quality can be obtained through the control and  curtailment
of both runoff and erosion shortly after pesticide application.

     As shown in Figure 40, the highest predicted alachlor concentration
during the three-year period occurred around the end of May to  early
June 1973.  Detailed simulated flow rate, sediment concentration and
concentrations of particulate, dissolved and total alachlor during this
period are shown in Figures 41, 42, 43 and 44.  As indicated by Figure 44,
on June 5, 1973, when the maximum alachlor concentration occurred,
sediment carries up to approximately 35% of the total alachlor  being
transported, while 65% was in a dissolved form, according to the model
prediction.

     Simulated longitudinal distributions of sediment concentration  and
concentrations of dissolved, particulate and total alachlor occurring on
June 4, 1973, are presented in Figures 45 and 46, respectively.

     Instream modeling demonstrates some important effects of  sediment
transport on pesticide migration:

 1.  Through adsorption of dissolved pesticides by sediment, immediate
     biological availability of the pesticide may be reduced.

 2.  Through deposition of contaminated sediment, pesticide concentrations
     in a water column will be reduced.

 3.  However, the contaminated bed sediment then becomes a long-term
     source of pollution through resuspension and desorption.

The effect of adsorption as described above is demonstrated in  Figures 44
and 46.  When the maximum alachlor concentration occurred near  the mouth
of Wolf Creek on June 5, 1973, approximately 35% of the total  alachlor was
being carried by sediment, while 65% was in a dissolved form that  is
generally subject to more immediate uptake by aquatic biota (Figure  44).
Near the Four Mile Creek Mouth, dissolved and particulate alachlor

                                     83
-------
consisted of 22% and 78% of the total, respectively.  Since most of
alachlor at the stream edge was in a dissolved form, sediment uptake in
the receiving water reduced the dissolved pesticide amount by 78%
(Figure 46).

     The effect of contaminated sediment deposition is clearly seen in the
variation of longitudinal alachlor concentrations in Wolf Creek
(Figure 46).  The percentage of particulate alachlor to total alachlor
steadily decreased from 59% to 8% with the downstream distance in Wolf
Creek.  The reduction of particulate concentration with the downstream
distance was due to the deposition of contaminated sediment before it
reached the mouth of Wolf Creek.  Another possible cause of particulate
contaminant reduction is the dilution of pesticide by clean Wolf Creek
water inducing desorption of pesticide from sediment to water.  Reduction
of total alachlor concentration with distance reflects the particulate
alachlor trend.

     The effect of sediment migration on contaminant distribution is
demonstrated in Figure 47, showing longitudinal particulate alachlor
distributions associated with the three sediment-size fractions and their
weighted average in the top bed layer after the 3-year simulation.  Since
it was assumed that initially there was no alachlor in the stream bed, the
accumulation of alachlor in the bed must have occurred during the 3-year
period through deposition and resuspension of contaminated sediment and
direct adsorption and desorption with overlying water.  Consequently, even
if the use of alachlor is terminated, already contaminated bed sediment
will continue to introduce alachlor back to the water column.

     Although predicted concentrations of sediment and pesticide obtained
by SERATRA seem reasonable, no detailed examination of these results was
not possible due to no measured data available.

     The dissolved pesticide distribution near the mouth of Wolf Creek
River Kilometer 5 shown in Figure 39 was then statistically analyzed by
FRANCO for the pesticide risk assessment.

STATISTICAL ANALYSIS AND RISK ASSESSMENT

Case Study of Alachlor

     Alachlor was assessed for its known lethal and sublethal effects to
fish.  The following toxicity information was obtained for Lasso® MCB/C-9,
Lasso® E.C., and for alachlor (100% active ingredient).  Lasso®(MCB/C-9)
was used for the risk assessment process. . The Lasso® E.G. formulation was
included to help formulate information on channel catfish.  The increased
toxicity of Lasso® over alachlor demonstrates the probable effects of sur-
factant synergism.

     With the exception of fathead minnow data which are from flow-through
measured experimentation, the results below are from static bioassays of
nominal (calculated) concentrations.  Optional standardization to flow-
through, measured concentration for LC50 values is shown below:

                                     84
-------
             ixicr6 pr
         en
        LU
        CJ
        to
        LU
        Q.
             IxlO'7
        1   1*10
-8
             1x10"
                                           	1
MEAN PARTI CULATE ALACHLOR
PARTI CULATE ALACHLOR WITH SAND
PARTICULATE ALACHLOR WITH
 SILT
PARTICULATE ALACHLOR WITH
 CLAY
                   	r
                  0       10     20      30     40      50      60

                          DISTANCE ABOVE THE MOUTH OF WOLF CREEK, km


Figure  47.   Variations of simulated particulate  alachlor  in  the top bed
             layer  accumulated  during the  three-year simulation  period.
                                       85
-------
     Static value x 0.71 approximates flow through concentration
     Nominal value x 0.77 approximates measured concentration
     Bluegill 96-hr LC50:  6.2 x 0.71 (static) x 0.77 (nominal) = 3.39
     Rainbow 96-hr LC50:  3.7 x 0.71 (static) x 0.77 (nominal) = 2.02

     Thus, the acute numbers standardize to roughly half their original
concentration values.  For lack of direct evidence indicating that these
specific values are more correct than the original concentration levels,
the original set was employed to continue the risk assessment.  However,
values obtained by standardization may more closely approximate actual
toxic concentration.  This emphasizes the necessity of accurate testing
techniques and recommends the use of flow-through bioassays of measured
concentration for best results.
                      LC50 in mg/1
                 24-hr   48-hr   96-hr
                                      Source
Lasso® (MCB/C-9)
  bluegill
  rainbow

Lasso® E.G.
  channel
  catfish
  bluegill

  rainbow
Alachlor
  bluegill
  rainbow
  fathead
  minnow
   16
    9.6
10
 7.8
   13
    9.2

    8.7
 6.4
 3.5

 5.8
6.2
3.7
                    6.5
                   13.4

                    2.3
2.8
1.8
Monsanto Agricultural Product Co.
Monsanto Agricultural Product Co.
               Monsanto Agricultural Product Co.
               Weed Science Society of America
               1979
               Weed Science Society of America
               1979
Monsanto Agricultural Product Co.
Monsanto Agricultural Product Co.
4.4   (192-hr LC50-2.5) Call et al. 1979
     The objective is to evaluate La'sso's® (MCB/C-9) toxicity for
bluegill, rainbow, and channel catfish.  Sufficient information is
provided for only the first two species, so results from Lasso® E.G. and
alachlor were used to derive a 96-hr LC50 value for catfish.  One would
expect the ratios of (MCB/C-9) to Lasso®E.G. data to be similar for
rainbow and bluegill.  This ratio could then be used to define  a value for
catfish.  Actually, however, there is a large difference in these ratios,
possibly due to the different effects of the emulsion formation on  the two
fish species.
Bluegill
                             = 0.46
     Rainbow
                          = 1.61
     Use was made of both ratios to calculate a range which may include
the true value for catfish.

     Catfish 6.5 x 0.46 = 3.0 and 6.5 x 1.61 = 10.5

                                     86
-------
     Assuming the 96-hr LC50  lies  between  3.0  and  10.5 mg/1,  a  geometric
mean, based on this range, for convenient  handling is calculated  as:
     Catfish 96-hr LC50:      so  x  10.5  =  5.6
     The next step is the calculation of 24- and  48-hr  LC50  values  for
catfish to Lasso® (MCB/C-9).  As  catfish are not  closely  related  to blue-
gill or rainbow,  the slopes from  24-, 48-, and  96-hr  LC50 data  of bluegill
and rainbow were  used to derive concentration ranges  for  catfish.
               Slopes
          Bluegill                         Rainbow
          6-2 - 16 mg/1 =     13g           3.7  -  9.6  =  _
            96 -  24 hr     u' JD            96  -  24
          6.2 - 10 _  0 07g                3.7 -  7.8 _  n
           96 - 48    U'079                 96 -  48	°'
     The ranges generated will be used to approximate the  24 and 48-hr
LC50 values for catfish.
     96-hr LC50 = 5.6 mg/1

       °"
     95.24
             = -0-082 to -0.136      '      = -0.082 to -0.136
Solving the above equations for X yields
     24-hr LC50 = 8.9 to 12.8 mg/1
     24-hr LC50 = 16.4 to 20.3 mg/1
Combining these values leads
     24-hr LC50 range = 8.9 to 20 mg/1
             = -0.079 to -0.085    °'5 7X = 0.079 to 0.085
Solving the above equations for X yields
           48-hr LC50 = 6.8 to 7.1 mg/1
           48-hr LC50 = 14.3 to 14.6 mg/1
Combining these results leads
     48-hr LC50 range = 6.8 to 14.6 mg/1
                                     87
-------
     The geometric mean converts these ranges to points.
          24-hr LC50: V8.9 x 20.3 = 13.4 mg/1
          48-hr LC50: \6.8 x 14.6 = 10.0 mg/1

     Standardization factors based on the 96-hr LC50 geometric mean of
5.6 mg/1 would give these results.  Slopes associated with these values
are dissimilar to those experimentally derived for bluegill and rainbow.

     5.6 x 1.52 = 24-hr LC50 of 8.5

     5.6 x 1.23 = 48-hr LC50 of 6.9

     The acute data considered most appropriate from the steps above are
included here in mg/1.  The symbol (~) above a number differentiates it as
an estimated (nonexperimental) value.

                        24-hr LC50    48-hr LC50   96-hr LC50

          Lasso (MCB/C9)

            bluegill        16            10           6.2

            rainbow          9.6           7.8         3.7

            catfish         13?4          10^0         5?6


     No chronic toxicity values were currently available for Lasso®
(MCB/C-9).  However, recent results of fathead exposure to alachlor show
the no-effect level between 0.52 and 1.0 mg/1 (Call et al. 1979).   Its
application factor range is 0.12 to 0.23.  Assuming the application factor
for Lasso® (MCB/C-9) is the same as for alachlor, the MATC factors  can be
calculated for the other fish species.

                       96-hr LC50                   MATC Range

          bluegill         6.2    x 0.12 to 0.23 = 0.74 to 1.43
          rainbow          3.7    x 0.12 to 0.12 = 0.74 to 0.85
          catfish          5.6    x 0.12 to 0.23 = 0.67 to 1.29

     Had an experimental MATC value for alachlor been unavailable,  selec-
tion of an arbitrary no-effect level would have been necessary,,  This pro-
cedure  is also given, and comparisons between the results of both methods
can be examined.  Since Lasso® is not known to be either persistent or
cumulative in its effects under normal conditions, the selected arbitrary
application factors are 0.05 for a 24-hour average and 0.1 for a temporary
level.  Use of the 24-hour average should give a conservative estimate of
the MATC value:

         96 hr-LC50 x AF = MATC
     bluegill 6.2 x 0.05 = 0.31
      rainbow 3.7 x 0.05 = 0.19
      catfish 5.6 x 0.05 = 0.28
                                     88
-------
In this case the arbitrary MATC  values  produce  a  lower,  more  conservative
measurement of the no-effect  level.

     The rainbow trout species (Salmo gairdneri)  was  chosen for  the
methodology illustration with the following  curves  using LC50 data and  the
MATC value.

 1.  The SERATRA cutoff.

 2.  The MATC line (see Figure 48).

 3.  The LC50-MATC curve (see Figure 49).

     The six pairs of points used for each curve  are  listed below.   Con-
centrations are given as kg/m3 (= 1000  ppm).  While six  points were  used
to define such curves, some could have  been  defined with fewer points.

 1.  Curve 1 uses the concentration of  1.0 x 10~9 at  each of  the time
     steps which include 0.5, 24, 48, 96, 168,  and  336 hours.

 2.  Curve 2 uses the concentration of  1.9 x 10~^ at  each time step
     including 0.5, 24, 48, 96,  168, and 336 hours.

 3.  Curve 3 uses the following  (duration, concentration) pairs:   (0.5,
     0.0096), (24, 0.0096), (48, 0.0078), (96,  0.0037),  (96,  0.00019),
     (336, 0.00019).

     The mouth of Wolf Creek has been used for  modeling  the concentration-
duration aspects of the FRANCO code.

     The Exceedance Summary for  alachlor includes several parameters for
each curve.  In this case the summary shows  that  the  duration-concentra-
tion levels of alachlor were insufficient to register as an event.   No
duration exceeded even the MATC function, and therefore,  global  exceedance
is 0%.  Investigation of the data generated  under each category  confirms
that the minimum given value for alachlor, its  estimated MATC concentration
of 1.9 x 10"4 kg/m3 (or 0.19 ppm), is not exceeded  at any time (see
Figure 39).

     Thus, it is concluded that dissolved concentrations of alachlor near
the mouth of Wolf Creek is less than the MATC for the three years modeled
and should present no direct threat of  toxicity to  the rainbow trout or
other species with a MATC greater than  0.19  ppm,  assuming no  unusual cir-
cumstances increasing the effective concentration or  effective toxic level.

Case Study of Tpxaphene

     Alachlor was chosen in the modeling study  because of its available
field data.  The computer evaluation demonstrated that alachlor  should  not
have caused damage to rainbow trout.  To more clearly illustrate possible
applications of the model, toxaphene was chosen as  a  second pesticide for
the risk assessment.

                                     89
-------
ppm


 .6

 .4
  ppm
  10
         24    48    72
96
Hours
2668
          Figure 48.  MATC line for alachlor.
           I
          24    48    72    96
                            Hours
          Figure  49.   LC50 -  MATC curve for Lasso®.

                          90
                        2668
-------
     Toxaphene, a chlorinated hydrocarbon,  is  very persistent  in  soil  and
water and may be bioconcentrated  to  a  high  degree.   Its  addition  to lakes
has been shown to cause toxicity  for ten  years.   Its  persistence  in soils
has been cited as 4  to 16 years.   Its  solubility of 1.5  ppm and  its esti-
mated Kd value of 5  x 10^ differ  considerably  from those of alachlor.
One would expect toxaphene to be  strongly sorbed to the  sediments,  raising
interest in modeling toxaphene  in sediments  as well  as the  dissolved state.

     The risk assessment procedure for toxaphene is based on the  same
stream concentrations as generated for alachlor.  This is done only as an
example and is not meant to  imply actual  concentration levels  for toxa-
phene.  The physical and chemical  properties of  toxaphene are  considerably
different from alachlor.

     Toxicity data used here are  from  pages  B-27 and  B-28 in the  Appen-
dix B.  The following species have been selected to illustrate the  assess-
ment procedure for toxaphene.   Again,  the values are  expressed as mg/1.

     Without knowledge of the technique of  each  bioassay, one  cannot make
a logical decision on the necessity  for standardization  of  the values.
For the sake of this example, it  was assumed that the following numbers
represent actual toxicity values.

                              24-hr  LC50    48-hr LC50    96-hr LC50


          Coho Salmon           0.0130        0.0105        0.0094

          Cffinook Salmon        0.0079        0.0033        0.0025

          Rainbow Trout         0.0115        0.0084        0.0084

          Brown Trout                                       0.003

          Carp                                              0.004

          Fathead Minnow                                    0.014

          Channel Catfish                                   0.013

          Guppy                                             0.020
          Bluegill Sunfish                                  0.0035
          Largemouth Bass                                   0.002


     The minimum three data points are available in Appendix B for  only
three species.  The rest must be  derived  from  these or from toxicity data
for related pesticides.

     The brown trout is related by genus  to  the  rainbow  trout  so  the slope
of the 24-96 and 48-96 hour segments from the  rainbow were  used to  esti-
mate the 24- and 48-hr LC50 values for the  brown trout.


               rainbow "•«% \ °j0115 .  -4.3  x  IP'5


                                     91
-------
                       0.0084 - 0.0084 _ n
                           96 - 48U

               brown   0.0.03 -4X B .4.3 x 10-5


                       24-hr LC50 = 0.0061 mg/1
                       0.003 - X _ n
                        96-48


                       48-hr LC50 = 0.003 mg/1


     None of the other species other than the salmon are related to the
trout except by class, so other steps are used to estimate the 24- and
48-hr LC50 values where they are missing.

     The use of the geometric mean of the slopes will be demonstrated to
approximate the unknowns.  There is no way of knowing whether values cal-
culated in this manner fall within the range provided by the other slopes.
However, the chances are improved with a larger number of experimental
values already established for each time interval.  The use of estimated
values from previous steps introduces more probability for error and should
be avoided.  In this example, the experimental values obtained for coho,
Chinook, and rainbow are used.

                                24 - 96-hr Slope    48 - 96-hr Slope

              coho salmon         -5.0 x 10"5         -2.3 x 10"5

              Chinook salmon      -7.5 x 10"          -1.7 x 10~
              rainbow trout       -4.3 x 10"5          0

                Range             -7.5 x 10"5         -2.3 x 10"5
                                to -4.3 x 10"5           to  0.0

                Geometric mean    -5.44 x 10"5        -1.58 x 10"5
                                                    (as 3V-2.3 x 1.7 x 10"10)


     Applying these values to the fish with incomplete data sets results
in the following ranges and geometric means:

     Example:

              (	
               96 - 24
     carp     °^04 :,X = -7.5 x 10"5 to -4.3 x  10"5
                              Range                    Geometric Mean

                      24-hr LC50 = 0.0071 to 0.0094         0.0079
                                      92
-------
Estimated Geometric Means as mg/1:
                          24-hr LC50    48-hr LC50   96-hr LC50
carp 0.008 0.005
fathead minnow 0.018 0.015
channel catfish 0.017 0.014
guppy 0.024 0.021
bluegill sunfish 0.0077 0.0040
largemouth bass 0.006 0.002
0.004
0.014
0.013
0.020
0.0035
0.002
The values above can be contrasted with those calculated using 24,
48, and 96-hr standardization factors:
96-hr LC50 x 1.52 = 24-hr LC50
96-hr LC50 x 1.23 = 48-hr LC50
24-hr LC50 48-hr LC50
carp 0.006 0.005
fathead minnow 0.021 0.017
channel catfish 0.020 0.016
guppy 0.030 0.025
bluegill sunfish 0.0053 0.0043
largemouth bass 0.003 0.0025


96-hr LC50
0.004
0.014
0.013
0.020
0.0035
0.002
In this case the two methods produced similar results. Because there
are some experimental data sets from which slopes can be derived, the
estimated geometric mean values were used as the most likely to represent
the actual values. This yields the following toxaphene acute toxicity
data as mg/1 for use with the probability model:
24-hr LC50 48-hr LC50
coho salmon 0.0130 0.0105
Chinook salmon 0.0079 0.0033
rainbow trout 0.0115 0.0084
brown trout 0.006 0.003
carp 0.008 0.005
fathead minnow 0.018 0.015
channel catfish 0.017 O.ol4
guppy 0.024 0.021
bluegill sunfish 0.0077 O.ot)40
largemouth bass 0.006 0.002
96-hr LC50
0.0094
0.0025
0.0084
0.003
0.004
0.014
0.013
0.020
0.0035
0.002
   The symbol (~) indicates estimated values.
                                93
-------
     Chronic toxicity values for toxaphene stated as MATC ranges  have  been
located for the fathead minnow and channel catfish.

          fathead minnow      0.000025 < MATC < 0.000054
          channel catfish     0.000049 < MATC < 0.000072

     Application factor ranges are calculated below from MATC  arid 96-hr
LC50 data.

     MATC range/96-hr LC50 = AF range
     fathead minnow      0.000025 to 0.000054/0.014 = AF
                              AF = 0.0018 - 0.0038
     channel catfish     0.000049 to 0.000072/0.013 = AF
                              AF = 0.0038 - 0.0055

     Carp and1 fathead minnows are related at the family level  so  the appli-
cation factor derived for the fathead minnow will be applied to the  96-hour
LC50 point for carp.

                       0.004  x 0.0018 to  0.0038 = MATC
                        0.0000072 < MATC < 0.0000152

     For the unrelated species the high and low values from fathead  minnow
and channel catfish can be used to set a range.  The geometric mean  can  be
used to convert the ranges to points.

     AF range = 0.0018 - 0.0055

     Toxaphene is known to be very persistent and would require the  use  of
the more conservative arbitrary application factor of 0.01  for a  24  hour
average if no experimental data were available.

     Geometric means, as well as MATC experimental and derived ranges, have
been included in this list of chronic toxaphene toxicity as mg/1.

                      	< MATC Range <	   Geometric Mean

     coho salmon
     chinook salmon

     rainbow trout
     brown trout

     carp

     fathead minnow

     channel catfish

     guppy

     bluegill
                                      94
1.7 x 10"5
0.45 x 10"5
1.5 x 10"5
0.54 x 10'5
0.72 x 10"5
2.5 x 10"5
4.9 x 10"5
3.6 x 10"5
6.3 x 10"5
3.6 x 10"5
5.2 x 10"5
1.4 x 10"5
4.6 x 10"5
1.7 x 10"5
1.5 x 10"5
5.4 x 10"5
7.2 x 10"5
11.1 x 10"5
1.9 x 10'5
1.1 x 10"5
3.0 x 10"5
0.79 x 10"5
2.6 x 10"5
0.96 x 10"5
1.3 x 10 "5
3.7 x 10"5
5.9 x 10 "5
6.3 x 10"5
3.5 x 10"5
2.0 x 10"5
-------
     Lethal and sublethal data was used to assess  environmental  damage  to
rainbow trout (Salmo gairdneri) from toxaphene exposure.   Because  of  toxa-
phene's extreme toxicity to fish, seven curves were  chosen,  in  addition to
the SERATRA cutoff, to demonstrate the kind of additional  information
which can be gained.  All are illustrated in Figures 50 through  57.

 1.  Curve A is the SERATRA cutoff value.

 2.  Curve B is the MATC for rainbow trout.

 3.  Curve C is the least conservative lethality curve.

 4.  Curve D is the most conservative curve designated as  LC50-MATC.

 5.  Curve E is intermediate between curves C and  D.

 6.  Curve F considered the results of the LC50-MATC line  segments after
     24 hours.

 7.  Curve G generates results exceeding the LC50-MATC lines  beginning  at
     48 hours.

 8.  Curve H describes exceedance of the MATC starting at  96  hours.

     The six pairs of points used for each curve are listed with duration
in hours and concentrations as Kg/nH.

 1.  Curve A uses the concentration of 1.0 x 10~9  Kg/m3 at each  of the
     following durations:  0.5, 24, 48, 96, 168, and 336 hours.

 2.  Curve B uses the most conservative MATC endpoint for  rainbow  trout of
     1.5 x lO'8 Kg/m3 .  Time durations are again  0.5, 24, 48,  96, 168,
     and 336 hours.

 3.  Curve C includes the following sets of points:
     (24, 1.15 x 10-4), (24, 1.15 x 10-5),
     (48, 8.4 x 10-6), (96, 8.4 x 10-6),
     (168, 8.4 x.10-6), and (336, 8.4 x lO'6).

 4.  Curve D is described by the following points:
     (0.5, 1.15 x 10-5), (24, 1.15 x 1Q-5),
     (48, 8.4 x 10-6), (96, 8.4 x 10-6),
     (96, 1.5 x 10-8), and (336, 1.5 x lO'8).

 5.  Curve E has points at (0.5, 1.15 x 10'5),
     (24, 1.15 x 10-5), (48> 8>4 x 10-6)f
     (96, 8.4 x 10-6), (168, 8.4 x 10'6), and
     (336, 8.4 x ID"6).

 6.  Curve F includes the following points: (24, 1.15 x 10"4),
     (24, 1.15 x ID'5), (48> 9.4 x 10-6)f (95, 9.4 x  uj-6)
     (96, 1.15 x 10-8), and (336, 1.5 x 10"8).

                                     95
-------
 Kg/nf
1x10
    -9
             I	I
I
             24     48     72    96
                       Hours
          2668
  Figure  50.   Curve A - SERATRA Cutoff value.
 Kg/rrT

2x10~£

1x10~£
                          I
 I
             24    48    72    96
                       Hours

        Figure 51.  Curve B - MATC line.
           2668
                       96
-------
     Kg/nr


        ~6
   12x10
   10x10
        ~6
    8xlO
        "6
    6x1 O
        "6
          .<
                 24    48     72     96

                           Hours
                                             2668
Figure 52.  Curve C - Lethality curve least likely to
            indiciate mortality.
      Kg/m°

         -6
    12x10
    10x10
         -6
     8x10
         -6
   1.5x10
         -8
                             j      i
                                              2668
                  24    48     72     96

                           Hours

Figure 53.  Curve D - LC50  -  MATC conservative curve.
                          97
-------
    Kg/nT

       -6
  12x10
  10x10
       -6
   8x10
       -6
   6x10
       -6
                 I
                    I
I
                   I
24
     96
                                             2668
                    48     72

                        Hours

Figure 54.  Curve E - LC50 - alternative  curve.
Kg/m3
12xlO"6
lOxlO"6
8x1 O"6
1.5xlO"8




-
- \
:
i i i i . ,
24 48 72 96 26
Hours
Figure 55.  Curve F - LC50 - MATC curve for durations
            exceeding 24 hours.
                         98
-------
Kg/m3
10xlO~6
8xlO"6
1.5xlO"8
0

-
I


i
1
1
i
I
1
i
i i / / 	 l
24 48 72 96 26
Hours
 Figure 56.  Curve G - LC50 - MATC curve for durations
             exceeding 48 hours.
Kg/m°
8x1 O"6
**
5xlO~8
n

-
^



i i i i //
                   24    48    72    96

                             Hours
2668
Figure 57.  Curve H - MATC exceedance for durations greater
            than 96 hours.
                           99
-------
 7.  Curve G has the following points:   (48,  1.15 x  10'4),
     (48, 1.15 x 10-5), (48, 8.4 x 10'6),  (96, 8.4 x lO'6),
     (96, 1.5 x 10-8), and (336, 1.5 x lO'8).

 8.  Curve H is described by the following points:
     (96, 1.15 x ID"4), (96, 1.5 x lO'5),  (96, 8.4 x 1CT6),
     (96, 1.5 x ID'8), and (336, 1.5 x lO'8).

     The exceedance summary inumerates the following curve results for  the
mouth of Wolf Creek.

          Curve    Events    Total Duration    Global Exceedance

            A         9         4160   hr            16.23%

            B        12         3828                 14.93

            C         2          314                  1.23

            D         6         3461.5               13.50

            E         2          314                  1.23

            F         6         3461.5               13.50

            G         6         3461.5               13.50

            H         6         3461.5               13.50


 1.  Curve A - 16.2% of total time divided into nine events exceeded  the
     SERATRA cutoff value.

 2.  Curve B - Twelve events, comprising 14.9% of the modeling  time,
     exceeded the MATC designated line of  1.5 x lO'5 ppm.  Had  the high
     MATC endpoint of 4.6 x 10~5 ppm or  the  geometric mean value  of 2.6
     x 10~5 been selected to represent the MATC, the global exceedance
     value would have been reduced to 14.0%  and 14.5%,  respectively.

     Of the twelve events, six lasted for  more than  96  hours  and  accounted
     for 90.4% of the time exceeding the MATC.  The  longest event of
     1257.5 hours began at time-step 16496 and lasted through time-step
     19011.  Clearly, 1257.5 hours at concentrations above the  MATC sug-
     gests concern for sublethal effects.  Damage may largely depend  on
     life stages present during such long-term events.

 3.  Curve C - The lethal curve based on known LC50  values was  exceeded in
     two events lasting a total of 314 hours, which  was 1.2%  of the total
     modeling time.  A breakdown of the  event durations shows 196.5 hours
     for the first event and 117.5 hours for the second.  The assumption
     of no lethality prior to D suggests that, in spite of values exceed-
     ing GI at DI, no median fish kill occurs.

 4.  Curve D - The MATC incorporated with  the LC50 curve  is represented by
     six events lasting 3461.5 total hours for a global exceedance of
     13.5%.  Those events, represented by  curve B (lasting more than
     96 hours) and by curve C, are included  in this  percentage.
                                     100
-------
 5.  Curve E - This curve is equivalent to curve C  in this example.   This
     may seem surprising since two events had concentrations  greater  than
     1.5 x 10~5.  However,  these concentrations lasted for 96  and
     168 hours, which is long enough to be counted  in curve C.  The assump-
     tion that C0 = GI indicates that most rainbow  trout would  not survive
     the first 24 hours of exposure at concentration levels greater than
     GI-  Had concentrations greater than 1.15 x 10~5 been evident for
     durations less than 24 hours, the global exceedance for  curve E  would
     be greater than curve C by the area represented by C  > C\  and D  >  0]_.

 6.  Curve F - Results of curve F are identical to  curve D because concen-
     trations greater than GI at durations less than DI exceed  the
     24 hour minimum function time.

 7.  Curve G - This curve again shows the same results as  curve D.  Both
     events exceeding curve G are also exceeded by  curves  0 and F.

 8.  Curve H - Once more, the function represented  by C  >. 1.5 x 10~5  ppm
     at D .>94 hours is exceeded six times for a total of  13.5%.  Had any
     of the six events noted in curves F or G above lasted for  less than
     96 hours, the exceedance of curve H would be reduced  proportionally.

     A summary of the curves is shown in Figure 58.  Concentrations are
within safe limits 85.1% of the total modeling time.  Concentration-
duration values fell within the .potential lethal/sublethal (
-------
   Kg/nr
      -6
 12x10
 10x10
      -6
  8x10
      -6
  6x10"
  4x10
      -6
1.5x10
      -8
                               1.2%
                        13.7%
               I      I
                                85.1%
MATC
              24    48    72    96    120

                             Hours


        Figure 58.   Summary  of  toxaphene  curves.
      2668
                           102
-------
                                  SECTION 6

                     EVALUATION OF THE CMRA METHODOLOGY

     The CMRA Methodology was developed  and was applied  to  the  Four  Mile
and Wolf Creek area in Iowa to examine the applicability and  limitation of
this methodology.  The methodology is general enough to  evaluate  the migra-
tion and risk assessment as applied to a wide range of pesticides, agri-
cultural lands, receiving streams and aquatic biota.  For example, this
methodology could be used to select optimum pesticide application practices
including timing and doses.

     The application of the methodology  to to Four Mile  and Wolf  Creeks
revealed that in general, the CMRA Methodology is capable of  simulating
migration and fate of a pesticide and its associated risk to  aquatic biota.
However, accuracy of predicted results depends largely on the availability
of data.  The application exercise also  revealed that it  requires a  rather
extensive effort to evaluate pesticide movements and its  risks.   Further-
more, due to a general lack of data, calibration and verification of the
models, especially portions concerning pesticide transport, adsorption/
desorption, and degradation were not fully performed and  that further test-
ing of the methodology is required.  Also a FRANCO output of  probability
of acute and chronic  damages to aquatic biota must be supplemented  by
other pertinent information on pesticide characteristics, environmental
stress, life stages of biota, etc., to obtain more comprehensive  risk
assessment.  Applicability and limitations of each of the four  components
of the CMRA Methodology are discussed below.

OVERLAND PESTICIDE MODELING

     The ARM model was specifically developed as a tool  to  evaluate  the
quantity and water quality of runoff coming from agricultural areas,  as
well as the impacts of alternative management practices.  The model's
capability to continuously simulate nonpoint source pollution processes
makes it useful for evaluating both short- and long-term migration of
pesticides overland.  The application of ARM is limited  primarily by the
availability of data for model calibration and by the size  of an  area
which can be accurately modeled in a single simulation.  Given  the state-
of-the-art of modeling nonpoint source pollution from agriculture, ARM is
the most appropriate pesticide loading technique for this methodology.

INSTREAM PESTICIDE MODELING

     The unsteady, two-dimensional model, SERATRA, which  is used for  the
instream pesticide modeling, includes all important mechanisms  of both


                                     103
-------
dissolved and particulate pesticide transport phenomena, such as convec-
tion and dispersion of pesticides, interaction of pesticides with sediment
(adsorption and desorption of pesticides with sediment, and transport,
deposition and resuspension of particulate pesticides associated with
sediment), and chemical (oxidation, hydrolysis, and photolyses) and
biological degradation of pesticides and volatilization.  Since SERATRA  is
one of the few tested models capable of simulating both dissolved and par-
ticulate contaminants with sediment-contaminant interactions, it is well
suited to this methodo"1ogy.  SERATRA is applicable to the nonticlal rivers
and narrow lakes where vertical and longitudinal distributions are of
interest, but is not suitable to estuaries and coastal areas where lateral
and longitudinal distribution may be of concern.  Another important aspect
of this portion of the methodology is that SERATRA provides vertical and
longitudinal distributions of dissolved and particulate pesticides, as well
as the pesticide accumulation in the stream bed.  However, because of the
rather limited knowledge of pesticide toxicity, only cross-sectionally
averaged dissolved pesticide concentrations are used for the risk assess-
ment.  Additional data collection for pesticide toxicity may make it pos-
sible to take the full advantage of detailed simulation results of SERATRA
in the future.  Lack of field data needed for the model calibration is the
major difficulty apparent when SERATRA is applied to an actual study area.
Another limitation is the limited applicability of EXPLORE-I to very small
streams.  Furthermore, since SERATRA is a two-dimensional (longitudinal  and
vertical) model and EXPLORE-I is one-dimensional (possibly used as longitu-
dinal and vertical two-dimensional model), compatibility between SERATRA
and EXPLORE-I is less than ideal.

STATISTICAL ANALYSIS AND RISK ASSESSMENT

     The risk assessment procedure uses a finite number of parameters to
summarize the pesticide concentration within the watershed of interest and
to predict the probability of toxic effects to indigenous fish species.  Of
necessity, the risk assessment procedure must be simplistic compared to  the
environmental complexities found within the actual system.  The acceptance,
or at least acknowledgement, of certain assumptions used in this procedure
allows for the prediction of environmental hazards to fish based on those
assumptions.

     Several important limitations exist in the risk assessment portion  of
the methodology.  Some of the major ones are these:

 1.  Laboratory bioassays are assumed applicable to the prediction of
     field toxicities.

 2.  Synergistic effects of pesticides with river water quality parameters
     are  ignored for lack of information.

 3.  Bioconcentration and biomagnification properties with predator-prey
     relationships were not assessed.

 4.  Interpretation of sublethal effects based on exceedance of the MATC
     may  overestimate hazards.

                                     104
-------
 5.  Effects of  ingesting pesticides from the  suspended  fraction  and  from
     bottom sediments have not been addressed.

 6.  Avoidance behavior and migration  resulting from  pesticide  concentra-
     tion in the water needs  investigation.

 7.  Assessment  for resistant or  susceptible strains  requires  identifica-
     tion of such properties  for  modification  of risk  assessment  input
     data.

     The FRANCO  program is very flexible in  its risk  assessment aspects.
The output describes pesticide concentration as the number  of  times a given
value  (Ci,dj) is exceeded, the percentage of total time  (C-j,dj) exceeded,
and individual and total time durations when the concentration  for a  par-
ticular river segment exceeds (C-j,dj).  It also computes  the percentage
of total time the designated  function  has been exceeded.  The model has
been demonstrated using only  the  dissolved portion of  pesticide.  The pro-
gram is also capable of assessing the  pesticide concentration  in  the  sus-
pended particulate and bed sediment fractions.

     The CMRA Methodology was used specifically with  fish species to  graph
the LC50 curve and the MATC and measure its exceedance.   However, any other
measure of effects may be used instead, such as LC10  or  LC90 values.  Con-
fidence limits are especially good for indicating a probability range of
toxicity.  The curve may be partitioned into acute and chronic  sections or
described as short-term and long-term  effects  with time  boundaries indi-
cated by the methodology user to  best  fit the  situation.  The  selection of
input values may be changed in accordance with the needs.   However, the
user must be cautioned to make appropriate interpretations  based  on the
assumptions used for data input.

     In spite of the limitations  inherent in FRANCO,  there  are  real bene-
fits associated with the modeling procedure for predicting  the  extent of
environmental hazards.  It has defined some of the problems which may be
resolved in the future by one of  at least three possible methods.  Addi-
tion of new mathematical relationships may allow for modeling  parameters
which more closely fit field conditions.  Improvement  of  data  collection
systems to fit modeling needs may reduce present model uncertainties. The
use of more appropriate hazard indicators and  resultant  interpretations
will improve the predictive ability of the modeling system.

MODELING FLEXIBILITY

     Although, as discussed above, the CMRA Methodology  consists  of four
components, the models can be used in  different combinations depending on
the problems.  However, the statistical analysis and  risk assessment  based
on FRANCO must be used together.  Examples of  combined uses are as
follows:

 1.  If the migration and risk assessment of a toxic  contaminant  are  of
     concern only at a stream edge, FRANCO can be used to summarize ARM
     results.

                                     105
-------
2.  If a toxic contaminant is directly discharged to a receiving water
    body, only EXPLORE-I, SERATRA, and FRANCO are required to evaluate
    transport and risk assessment.

3.  If there are continuous measurements of toxic chemical concentrations
    available and only the risk assessment is needed, FRANCO alone can  be
    used.

4.  If the receiving water bodies are estuaries or coastal waters,
    SERATRA can be replaced by other appropriate sediment-contaminant
    models such as FLESCOT and FETRA (Onishi and Wise 1978).
                                    106
-------
                                  SECTION 7

                     SIMPLIFIED STATISTICAL METHODOLOGY
     Overland and instream pesticide concentrations were  statistically
examined to study whether pesticide concentration patterns  can  be  expressed
by statistical distributions.   In this case, gamma, normal  and  log-normal
distributions were selected for testing.  This task was conducted  to  deter-
mine if a simpler pesticide assessment methodology which  relies only  on
the statistical nature on precipitation, watershed characteristics, and
receiving water body characteristics, could be developed  to replace the
more detailed computer-model-based CMRA Methodology, as described  in
Section 4.

     The principal steps in the development of any simplified statistical
methodology are the statistical summarization of the system input  (precip-
itation), the statistical summarization of the system output (e.g., dis-
solved, particulate, and total  pesticide concentrations)  and a  simple pro-
cedure for predicting the output from the input.  The development  of  these
steps begins by first determining the information desired.   The stated
objective is the frequency of occurrence and duration of  given  pesticide
concentrations.  In Subsection  4.2 seven specific summary measures are
given to meet this objective.   One particular measure, Pj[C >.C-j], gives
the relative frequency of occurrence of simulated concentrations greater
than or equal to specified concentration levels.  This measure  was chosen
for the simplified statistical methodology, because it is one of the
simplest summaries of a concentration time series.

     The FRANCO program summarizes the time series of instream  dissolved,
particulate and total (sum of dissolved and particulate pesticides) pesti-
cide concentrations.  The series is characterized by periods when  pesti-
cide concentration is lower than the chosen SERATRA cutoff  value,  and other
periods when runoff events cause the concentration to increase  above  this
value.  The cutoff value is selected to represent the point below  which
the simulated concentration values are due to limitations on the numerical
accuracy of the model and the computer.  Hence, time periods below this
concentration are considered to be zero.  As a result the relative fre-
quency of pesticide concentration consists of two parts:  A discrete  part
at the SERATRA cutoff value with an associated frequency  that estimates
the percent of the time the concentration is considered to  be zero and a
continuous part that summarizes the relative frequency of concentrations
above the cutoff value.

     Let GI equal zero, C2 equal the cutoff value and the remaining C-j
chosen such that C3 < C4 < . .  . < Cm.  Using the notation  introduced in

                                     107
-------
Section 4.2, this choice results in Pj[C  >C]J = 1 and Pj [C ^62! equal
to the relative frequency the simulated concentra- tion was greater than
or equal to the cutoff value.

Define

     6(C) = pF(C) + (l-p)I(C)

where   p = Pj [C >_ 63] is the relative frequency of concentrations
            occurring above the cutoff value, I(C) is a discrete cumulative
            frequency function with probability one at the cuttoff value,
            and F(C) is a continuous distribution function.

     The investigation of the feasibility of a simplified statistical
methodology is based on analyzing the cumulative distribution function
F(x) resulting from the dissolved, particulate and total pesticide concen-
tration time series from SERATRA for instream and ARM for stream-edge.
The same model can also be applied to the stream discharge and overland
runoff.  The latter tow can then be compared to an analysis of precipita-
tion data.  These analyses form the basis for discussing the feasibility
of a simplified statistical methodology.  Each of these analyses are  indi-
vidually discussed, followed by a discussion of their interrelationships.
The dissolved, particulate and total pesticide concentrations (in kg/rtH)
simulated by SERATRA near the mouth of Wolf Creek (River Kilometer 5)  is
statistically summarized below.

     During the three-year period starting June 1971 the dissolved pesti-
cide concentration was greater than the cutoff value of 1.0 x 1Q~9 for
16.23 percent of the time.  Hence p is estimated as 16.23.  The continuous
part F(C) of the cumulative distribution function 6(C) is well represented
by the log-normal distribution.  A probability plot construction from the
cumulative frequency concentration data illustrates this fit in Figure 59.
Similar probability plots using the normal and gamma distributions as
alternative choices did not result in the expected straight line plot
characteristic when the data fits a particular probability distribution.
The log-normal probability density function  is:


               ,    - i (In X - u)2/a2
     f (X) = - i— e  *                          X  > o.
            Xa
Stated  in this form, the parameters  u  and o^  are  estimated  by:
     ,2    ^  (In X  -u)2
     a  = J  1       '
              n -  1
                                     108
-------
    -8.1
     -11

  
-------
     The statistical distribution summary for the dissolved pesticide con-
centration is completed by the program FRANCO.  The distribution parameter
estimates are obtained from all concentrations greater than or equal to
the SERATRA cutoff value.  Because of the number of time steps involved,
these estimates are computed by summarizing the data with a flow through
technique that does not require all the data to be present at once.  In
constructing the probability plots ,a modification of the standard proce-
dure, such as described by Hann (1977), was necessary.  It was riot feasi-
ble for all data points to be plotted individually for the probability
plot.  To overcome this difficulty the data were accumulated as an empiri-
cal cumulative distribution function with up to 50 intervals used for the
concentration.  The probability plot was constructed using those values.
The effect on the plot is minimal if these concentration values are appro-
priately selected to result in approximately an equal number of time steps
in each interval.  The interpretation of the plot is not affected.

     The dissolved pesticide concentration at Wolf Creek River Kilometer 5
is adequately summarized by the mixed distribution:

     G(C) = 0.1623LN(C;-14.83,2.26) + .8377I(C)

where


     nn _ f 1   C < 1.0 x IQ'l
     i((j)   10   C > 1.0 x 10"y
and LN(C;u,a) denotes the log-normal distribution with parameters  u and  a.
A severe limitation of this summary  is the total loss of  all  information
concerning the actual time sequence  of the concentrations.

     A similar analyses of computed  results of SERATRA simulation  at Four
Mile Creek River Kilometer 5 was completed.  A mixed distribution  using  the
log-normal distribution for the continuous part again provided  the best  fit
when compared to the normal and gamma distributions as alternatives.  The
probability plot in Figure 60 illustrates the general conformance  to a
straight line.  The corresponding plot for the gamma distribution  (Fig-
ure 61) is not a straight line, indicating the gamma is  not  as  good a fit
for this data set.  The distribution with estimated parameters  is:

     G(C) = 0.1472 LN(C;-13.55,2.42) + 0.8528I(C)

where the terms are as previously defined.  These parameter  values result
in an estimated mean value of 2.44 x 10~5 with a standard deviation of
4.55 x ID'4.

     The summary statistics at Four  Mile Creek River Kilometer  5  and Wolf
Creek River Kilometer 5 show the effect of the instream  model embodied in
SERATRA.  First, the pesticide concentration  is above the cutoff  value
longer near the Wolf Creek mouth  (16.23%) than near the  mouth of  Four Mile
Creek (14.72%).  Second, the average concentration downstream in  Wolf Creek
is approximately five times smaller  than that in Four Mile Creek.   If the
                                     110
-------
    -7.8
     -10
  CO
  o
  	I
  <
  o
  o
     -16
     -18
     -21
          -20
                -18
 -15          -12

LOG (concentration)
-10
-7.5
Figure 60.
                  Log-normal probability plot  for  dissolved  pesticide
                  concentration at Four Mile Creek River  kilometer  5
                  (u = -13.55, a = 2.42)
comparison is made in log concentrations where the estimated  distribution
is normal, then it appears that the difference is mainly a  shift  in  the
mean value with the log variability approximately the  same  between these
two locations.

     The pesticide attached to particulates was  also summarized  using  the
mixed distribution analysis just described.  The distributions of the  simu-
lated results for particulate pesticides (in kg/m3) from ARM, at  Four
Mile Creek River Kilometer 5 and at Wolf Creek River Kilometer 5  were  again
a mixed distribution with the log-normal distribution  as the  continuous
form.  At stream-edge,

     6(C) = 0.0017 LN (C;-17.27,1.92) + 0.9983 I(C)

at Four Mile Creek River Kilometer 5

     G(C) = 0.1367 LN (C;-17.34,2.88) + 0.8633 I(C)

                                     111
-------
  10x10
   8x10
       -5
c/o
= 6x10
a
-------
     The distribution analysis was also applied to the dissolved  pesticide
concentration at stream-edge as simulated by the ARM model.  The  mixed  dis-
tribution model with the log-normal as the continuous part again  was  an
adequate fit.  In this case the gamma distribution was a close competitor.
The log-normal was selected to facilitiate the comparison with the SERATRA
results.  The estimated distribution is:

     G(C) = 0.0707 LN(C; -10.66, 2.45) + 0.9293 I(C)

     T/n     I I   C < 2 x TO'9
     ULJ     1 0   C > 2 x TO'9.


In comparison with the instream results, the stream edge concentrations
above the cutoff value were less than half as long (7.07%) and the concen-
tration distribution was shifted to higher concentrations with no differ-
ence in the variability in log units.  The average concentration  based  on
the log-normal distribution was 4.72 x 10~4 with a standard  deviation of
9.48 x 10'3.  The general effect of SERATRA on the pesticide distribu-
tion is to lengthen the time the concentration is above the  cutoff value
and to reduce the average concentration level.

     The statistical summary of the pesticide concentrations from ARM and
SERATRA has demonstrated that it is possible to describe the distribution
with a mixed probability distribution with the continuous part being  a
log-normal distribution.  The estimated parameters for this  mixed distri-
bution reflect the anticipated pesticide concentration changes expected
when going from stream-edge, to instream, and instream further down-
stream.  A severe limitation of this type of summary is the  loss  of all
the time sequence history of the concentrations.  In the present  work no
method was found to include duration of events in a distribution  analysis.

     The distribution analysis illustrates the feasibility of summarizing
the pesticide concentration distribution at stream-edge and  at selected
instream locations.  The summaries were shown to have the same mixed  dis-
tribution form with only the values of the parameters in the distribution
changing at each location.  To simplify the methodology, the relationship
among the parameters across locations must be determined.  The case study
illustrated the parameters changed in the direction intuitively expected.
However, it is not possible to determine a quantitative relationship.   A
large number of case studies at different sites, each simulated under a
variety of conditions, would be necessary to determine a quantitative rela-
tionship.  The computational time required for each simulation may become
excessive when a large number of simulations must be completed.   A second
limitation is related to the calibration procedures presently required  for
both the ARM and SERATRA models.  If the calibration is unique to each  site
of application, then the simplified statistical methodology  must  be able to
account for this. The severity of the problem can not be determined until
more case studies are completed.

     Additional study on a simplified statistical methodology must con-
sider another question.  What is the minimal information required concern-
ing pesticide occurrence and duration for a useful risk assessment?   An
                                     113
-------
answer to the question along with a specific statistical summary procedure
must be available before detailed studies of a simplified approach can be
useful.  Experience in applying the present CMRA Methodology will be use-
ful in determining the information necessary for a risk assessment and in
formulating an effective statistical summary for the  information.
                                     114
-------
                                  REFERENCES

Adams, R. T. and F. M. Kurisu.  1976.  Simulation of Pesticide Movement on
Small Agricultural Watersheds.  EPA-600/3-76-077.  U.S. Environmental Pro-
tection Agency, Washington, DC.

Baca, R. 6., W. W. Waddel, C. R. Cole, A. Brandstetter and 0. B. Cearlock.
1973.  EXPLORE-I:  A River Basin Water Quality Model.  Submitted to U.S.
Environmental Protection Agency.
WA  99352.
Pacific Northwest Laboratory, Richland,
Baker, J. L., H. P. Johnson, M. A. Borcherding, and W. R. Payne.  1979.
"Nutrient and Pesticide Movement from Field to Stream:  A Field Study."  In
Best Management Practices for Agriculture and Silviculture, Proceedings of
the 1978 Cornell Agricultural Waste Management Conference.Ann Arbor
Science Publishers, Inc., Ann Arbor, MI.

Call, D., L. Brooke, and R. Kent.  1979.  Estimates of No-Effect Concen-
tration of Selected Pesticides in Freshwater Organisms.  Third Quarterly
Progress Report to EPA, 1 January through 31 March 1979.  U.S. EPA Grant
No. R-806196-01.  University of Wisconsin, Superior, WI.

Crawford, N. H. and A. S. Donigian.  1973.  Pesticide Transport and Runoff
Model for Agricultural Lands.  EPA-660-2-76-013, U.S. Environmental Protec-
tion Agency, Washington, DC.
Crawford, N. H. and R. K. Linsley.
Stanford Wasteshed Model IV.
  1966.
             Digital Simulation in Hydrology:
Department of Civil Engineering, Technical
Report No. 39, Stanford University, Stanford, California.

Dawson, 6. W., M. W. Stradley, A. J. Schuckrow.  1974.  Determination of
Harmful Quantities and Rates of Penalty for Hazardous Substances.  Pre-
pared for the U.S. Environmental Protective Agency by Pacific Northwest
Laboratory, Volume 3 - Appendices, Washington, DC.

Donigian, A. S. and N. H. Crawford.  1976.  Modeling Pesticides and
Nutrients on Agricultural Lands.  EPA-600/2-76-043, U.S. Environmental
Protection Agency, Washington, DC.

Donigian, A. S., 0. C. Beyerlein, H. H. Davis and N. H. Crawford.  1977.
Agricultural Runoff Management (ARM) Model Verision II:  REfinement and
Testing.EPA-600/3-77-098, U.S. Environmental Protection AGency,
Washington, DC.
                                     115
-------
Donigian Jr., A. S., and H. H. DAvis, Jr.  1977.  User's Manual for
Agricultural Runoff Management (ARM) Model.  EPA-600/3-78-080, U.S.
Environmental Protection Agency, Washington, DC.

Donigian, A. S. and H. HJ. Davis.  1978.  Agricultural Runoff Management
(ARM) Model User's Manual:  Versions I and IEEPA-600/3-78-080, U.S.
Environmental Protection Agency, Washington, DC.

Eaton, J.G.  1973.  "Recent Developments in the Use of Laboratory Bioassays
to Determine "safe" Levels of Toxicants for Fish."   In Bioassay Techniques
an Environmental Chemistry.  Ann Arbor Science Publishers, Inc., Ann Arbor,
MI.

Ellis, B. G., A. E. Erickson, A. R. Wolcott, M. Zabile and R. Leavitt.
1977.  Pesticide Runoff Losses from Small Watersheds in Great Lakes Basin.
EPA-600/3-77-112, U.S. Environmental Protection Agency, Washington, DC.

Environmental Protection Agency.  1972.  Effects of Pesticides in Water - A
Report to the States.  PBB222-320, Prepared for Office of Water Resources
Research by NTIS, U.S. Department of Commerce.

Environmental Protection Agency.  1978.  Kepone Mitigation Project Report.
Standards and Criteria, Office of Water and Hazardous Materials, U.S.
Environmental Protection Agency, Washington, DC.

Environmental Protection Agency.  1979.  "Water Quality Criteria."  Federal
Register, Vol. 44, No. 52, pp. 15296-15981, U.S. Environmental Protection
Agency, Washington, DC.

Fread, D. L.  1973.  "Technique for Implicit Dynamic Routing in Rivers with
Major Tributaries," Water Resources Research, Vol. 9, No. 4, pp. 918-926.

Frere, M. H., C. A. Onstad and H. N. Holtan.  1976.  Actmo, An Agricultural
Chemical Transport Model.  Agricultural Research Service, ARS-H-3, U.S.
Department of Agriculture, Washington, DC.

Guenzi, W. D. (Ed.)  1974.  Pesticides in Soil Water.  Soil Science Society
of America, Inc., Madison, WI.

Haan, C. T.  1977.  Statistical Methods in Hydrology.  Iowa State
University Press, Ames, IA.

Herbicides Handbook of the Weed Science Society of America.  1970.  Weed
Science Society of America, 2nd Edition, pp. 80-83, W. F. Humphrey Press,
Geneva, NY.

Jordan, G. L.  1978.  "Environmental FActors and Soil Relationships
Influencing the Activity of Acetauilidie Herbicides."  Ph.D. Dissertation,
University of Wisconsin, Madison, WI.
                                     116
-------
Laflen, J. M., J. L. Baker, R. 0. Hartwig, W. F. Buchele and H. P. Johnson.
1976.  "Soil and Water Loss from Conservation Tillage Systems."  Trans-
actions of the ASAE, 21/5:881-885.

Leendertse, J. J.  1970.  A Water Quality Simulation Model for Well Mixed
Estuaries and Coastal Seas'. Principles of Computation, Volume I.
RM-6230-RC, The Rand Corporation, Santa Monica, CA.

Leytham, K. M., and R. C. Johanson.  1979.  Watershed Erosion and Sediment
Transport Model.  Submitted to the Environmental Protection Agency by
Hydrocomp, Inc., Palo Alto, CA.

McElroy, A. D. et al.  1976.  Loading Functions for Assessment of Water
Pollution Nonpoint Sources.  Midwest Research Institute, (PB-253 325).

Norton, W. R., L. A. Roesner, D. E. Evenson and J. R. Monser.  1974.
"Computer Program Documentation for the Stream Quality Model, Qual-II."
Water Resources Engineers, Inc., Walnut Creek, CA.

Oak Ridge National Laboratory.  1978.  Proceedings of a Workshop on
Evaluation of Models Used for the Environmental Assessment of Radionuclide
Releases.CONF-770901, September 6-9, 1977, Gatlinburg, TN.

Olsen, A. R., and S. E. Wise.  1979.  Frequency Analysis of Pesticide
lys
* O •
Concentrations for Risk Assessment.  Submitted to U.S. Environmental
Protection Agency.  Battelle, Pacific Northwest Laboratories,
Richland, WA.

Onishi, Y., P. A. Johanson, R. G. Baca and E. L. Hilty.  1976.  Studies of
Columbia River Water Quality - Development of Mathematical Models for
Sediments and Radionuclide Transport Analysis.  BNWL-B-452.  Battelle,
Pacific Northwest Laboratories, Richland, WA.

Onishi, Y. and S. E. Wise.  1978.  "Mathematical Modeling of Sediment and
Contaminant Transport in the James River Estuary."  Proceedings of the 26th
Annual ASCE Hydraulic Division Specialty Conference on Verification of
Mathematical and Physical Models in Hydraulic Engineering, pp. 303-310,
College Park, MD, August 9-11, 1978.

Onishi, Y.  1979a.  "Evaluation of Hazardous Substance Transport Modeling
in Surface Waters."  Proceedings of EPA National Workshop on Verification
of Water Quality Models, West Point, NY. March 7-9, 1979.

Onishi, Y., D. L. Schreiber and R. B Codell.  1979b.  "Mathematical
Simulation of Sediment and Radionuclide Transport in the Clinch River,
Tennessee."  Proceedings of ACS/CSJ Chemical Congress, Honolulu, Hawaii,
April 1-6, 1979.  "Contaminants and Sediments," R. A. Baker (Ed.), Ann
Arbor Science Publishers, Inc., Ann Arbor, MI.
                                     117
-------
Onishi, Y. and S. E. Wise.  1979c.   Mathematical Model, SERATRA, for
Sediment and Contaminant Transport in Rivers and its Application to
Pesticide Transport in Four Mile and Wolf Creeks in Iowa.  Submitted to the
U.S. Environmental Protection Agency.Battelle, Pacific Northwest
Laboratories, Richland, WA.

Onishi, Y. and S. E. Wise.  1979d.   User's Manual for the Instream Sediment-
Contaminant Transport Model. SERATRA.  Submitted to the U.S. Environmental
Protection Agency.  Battelle, Pacific Northwest Laboratories, Richland,
WA.

Onishi, Y.   1979e. User's Manual for EXPLORE-I:  A River Basin Water
Quality Model (Hydraulic Module Only).Submitted to the U.S. Environmental
Protection Agency.  Battelle, Pacific Northwest Laboratories, Richland, WA.

Roesner, L. A., et al.  1976.  Agricultural Watershed Runoff Model for the
Iowa-Cedar River Basins.  Report for U.S. Environmental Protection Agency,
Systems Development Branch, Washington, DC.

Sprague, J. B.  1969.  "Measurement of Pollutant Toxicity to Fish:  Bioassay
Methods for Acute Toxicity."  Water Research, 3:793-821, Pergamon Press,
GREAT BRITAIN.

Stephan, C. E.  1977.  "Methods for Calculating an LC50."  In Aquatic
Toxicology and Hazard Evaluation Proceedings of the Annual Symposium on
Aquatic Toxicology (ASTM), Special  Technical Publication 634:65-83,
American Society for Testing and Material, Philadelphia, PA.

Stewart, B. A., D. A. Woolhiser, W. H. Wischmeier, 0. H. Caro, and
M. M. Frere.  1976.  Control of Water Pollution for Cropland:  Volume I -
An Overview.  Agricultural Research Service, U.S. Department of
Agriculture, Washington, DC.

Water Quality Criteria.  1972.  Committee on Water Quality Criteria,
EPA-Re-73-033.  U.S. Environmental  Protection Agency, Washington, DC.

Wauchope, R. D.  1978.  "The Pesticide Content of Surface Water Draining
from Agricultural Fields - A Review."  Journal of Environmental Quality,
Vol. 7, No. 4.

Wuhrmann, K. 1952.  "Concerning Some Principles of the Toxicology of Fish."
Bull. Cent. Gelge. Docum. Eaux.  (Fisheries Research Board of Canada
Translation Series, No. 243.)  No.  15:49.

Weed Science Society of America.  1979.  Herbicide Handbook.  Fourth
Edition.  Champaign, IL.
                                     118
-------
                                 APPENDIX A

                 PROCEDURE FOR CMRA METHODOLOGY  APPLICATION

     Step-by-step  instructions to  use  the CMRA Methodology will  be
described.

OVERLAND PESTICIDE MODELING

     The first step in the CMRA Methodology  is the  application  of ARM to
the selected study area.  The procedure to be followed  is  essentially the
same as the five major steps outlined  by Donigian and Davis  (1977).   These
steps are discussed briefly below  and  have been  slightly modified for use
in this study.

Step 1:  Data Collection and Analysis

     Once the pesticide, application practice and study area  have been
selected, the input data for ARM must  be collected  and  analyzed.  These
input or execution data basically  consist of meteorologic  data,  such  as
precipitation, potential evapotranspiration, maximum-minimum  air tempera-
ture, wind movement, dewpoint temperature and solar radiation.   If snow-
melt calculations are not required, only the first  two  types  of meteoro-
logic data are needed.  Because ARM is a continuous simulation model,  these
data are required for the entire duration of the simulation time period.
The length of this time period is  primarily  determined  by  statistical
considerations to be discussed later as well as  data availability and the
required computation time.

Step 2:  Preparation of Meteorologic Data and Model Input  Sequence

     This step cpnsists of the construction of input data  files for each
type of meteorologic data.  The procedures for constructing these files
and their input formats have been  described  in detail by Donigian and
Davis (1977).

Step 3:  Parameter Evaluation

     Several of the ARM model parameters can be  evaluated  directly without
going through  the calibration process.  The types of information required
to evaluate these parameters include topographic maps,  soil maps, hydro-
logic/meteorologic studies, water  quality studies, cropping pattern sur-
veys, and data on pesticide application rates and modes.   The specific
parameters which can be evaluated  in this manner and methods  of evaluation
have also been described by Donigian and Davis (1977).
                                    119
-------
Step 4:  Model Calibration and Verification

     Calibration is an iterative process of making computer runs and coef-
ficient adjustments until simulated results match observed results.  It  is
required for parameters which cannot be determined directly.  The calibra-
tion process for the ARM model has been refined over several years of
effort and is discussed in great detail by Donigian and Davis (1977).

Step 5:  Generation of Runoff and Edge-of-Stream Sediment and Pesticide
         Loadings

     Upon completion of the previous steps, the ARM model can be used to
generate the runoff and sediment and pesticide loading information required
as input to the instream modeling component of the methodology.  This step
basically involves obtaining historical meteorological data for a given
time period or the selection of data which represent a specific sequence  of
storm events.  Pesticide application times can be based on local conditions
or selected arbitrarily.  In the latter case, pesticide removal by runoff
and erosion can be maximized by selecting pesticide application times sev-
eral days to several weeks before major storm events.  Pesticide applica-
tion rates and modes and cropping patterns can be obtained from local
information; if the pesticide being evaluated is new, application rates
and modes can be obtained from the manufacturer.

     The information generated in this step of the methodology is a  time
history of each of the following:  runoff, sediment loading, pesticide
dissolved in the runoff and pesticide attached to the sediment.  Depending
on the size and configuration of the watershed, this information is  gener-
ated for each catchment individually or for the total watershed.  The
in-stream sediment and contaminant transport model, SERATRA  (Onishi  and
Wise 1979a,b), requires the pesticide attached to each size fraction of
sediment being modeled in the river system.  However, ARM only simulates
total sediment loading and total pesticide adsorbed to the sediment.
Therefore, the ARM simulation results have to be modified prior to their
input to SERATRA.

INSTREAM PESTICIDE MODELING

     Instream pesticide modeling requires two computer models.  EXPLORE-I
(Baca et al. 1973; Onishi 1979) is used to obtain one-dimensional, unsteady
distributions of discharge (or velocity) and depth  in a receiving water
body.  Computed runoff is input to EXPLORE-I as a tributary  contribution.
SERATRA is used to simulate pesticide migration by  solving both sediment
and pesticide transport.  SERATRA uses EXPLORE-I results and computed sedi-
ment and pesticide loading from overland as input data.  Simulation  results
of pesticide distributions in a receiving stream are then statistically
analyzed to be used for the pesticide risk assessment.  Because of the  lack
of toxicological knowledge, only dissolved pesticide distributions are  used
for the risk assessment.
                                      120
-------
Step  1:  Hydrodynamic Data Collection  and Analysis:

      Once the  study area  is selected,  the input  data  for  EXPLORE-I  must
be collected and analyzed.  These  data consist of  channel  geometry  and
the Manning's  coefficient of the stream.  If  there is a dam  in  the  study
area, obtain the operational characteristics  of  the dam to be  inputed to
EXPLORE-I.  Boundary conditions at the upstream  and downstream  ends and at
tributary confluences must be  specified.  These  boundary  conditions may be
time  varying inflows or water  surface  elevations during the  entire  simula-
tion  period.   Initial flow conditions  must  also  be specified.   The  simula-
tion  period is the same as that used in the overland  modeling  component.

Step  2:  Preparation of EXPLORE Input:

      Assemble  the input data outlined  in Step 1  along with the  runoff com-
puted with ARM.

Step  3:  EXPLORE-I Calibration and Verification:

      EXPLORE-I must then be calibrated to match measured  depth  and  velocity
distributions  by adjusting the Manning coefficient.   If more than one set
of measured data are available for different  discharges,  use the first  set
of observed data for calibration and the other set(s) of  data for model
verification.

Step  4:  Production Run for Hydrodynamic Simulation:

      Upon completion of above  steps, run the  EXPLORE-I model to obtain
depth and velocity (or discharge)  distributions.

Step  5:  Sediment and Pesticide Data Collection and Analysis

      Collect necessary input data  of sediment and  pesticide  characteris-
tics, as well  as initial and boundary  values  of sediment  and pesticide  con-
centrations.   Detailed information of  SERATRA formulation, data need and
modeling procedure is described in Onishi and Wise (1979a,b).   Based on the
analysis of data availability, decide  whether to treat sediment as  one  bulk
sediment, consequently one bulk particulate pesticide, or  to divide the
sediment into  three sediment size  fractions (or sediment  types), resulting
in three groups of particulate pesticide associated with  each size  frac-
tions (or type) of sediment.  The  SERATRA user must realize  that the dis-
crepancy between actual and simulated  sediment transport  rates, hence also
particulate pesticide transport rates  will probably be large if only a
single size fraction (or single type)  of sediment  is  considered.  The use
of three size fractions (or types) is  itself  a compromise.   If  sediments
are divided into three size fractions  (or sediment types), then bulk sedi-
ment  and pesticide loading computed by ARM must be separated into each  size
fraction (or sediment type), as described in  Subsection 4.3.2.  Care must
be taken for selection of pesticide distribution coefficients.
                                      121
-------
Step 6:  Preparation of SERATRA Input

     Using data prepared under Step 5, and results obtained by ARM and
EXPLORE-I, assemble the input data for SERATRA.

Step 7:  SERATRA Calibration and Verification

     With available field data, adjust model parameters, especially those
associated with erosion and deposition of cohesive sediments  (i.e., criti-
cal shear stresses and the erodibility coefficient) and dispersion coeffi-
cient, as model calibration.  If there are additional field data for other
flow conditions, use the sets of field data for model verification.

Step 8:  Production Run for Sediment-Pesticide Transport Simulation

     Upon completion of the previous steps, run SERATRA to obtain unsteady,
vertical and longitudinal distributions of sediment and pesticide (both
dissolved and particulate) concentrations.  Sediment and pesticide condi-
tions in the river bed will also be computed.  Cross-sectionally averaged
dissolved pesticide concentrations will then be used for the  statistical
analysis.

STATISTICAL ANALYSIS

     The CMRA Methodology procedures require the  summary of SERATRA output
results of dissolved pesticide concentrations.  Each of these areas are
described in the order of their occurrence in the methodology.

Step 1:  FRANCO (Olsen and Wise 1979) Analysis of SERATRA  Instream
         Pesticide Concentration

     The FRANCO program requires particular information to be available
for a meaningful analysis (Olsen and Wise, 1979).  The SERATRA model cut-
off pesticide concentration value and the maximum pesticide concentration
are needed.  A time series plot of the data is helpful.  An array of user
specified concentration values, a similar array of duration values and a
number of LC (lethal concentration) functions are needed to control the
summary of the simulated pesticide concentrations.  The LC functions are
specified to provide the information needed for the risk assessment.  The
selection of these functions  is discussed in the  following risk  assessment
steps.  The concentration and duration values are selected with  considera-
tion for both the risk assessment needs and the statistical requirements
for providing detail on the frequency of occurrence and duration of spe-
cific pesticide concentrations.

RISK ASSESSMENT

Step 1:  Lethality

     Data must be located for pertinent aquatic species  in response to
the pesticide under investigation.  These values  should be stated as LC50
                                     122
-------
(median lethal concentration) or TLm (median tolerance  limit) for a spe-
cific time interval such as 24-hr LC50.  The minimum number of points
required include the 24-, 48-, and 96-hr LC50 numbers,  for each pesticide
and species.  Include the confidence limits around each point and any
experimental conditions specifically indicated.   Incipient LC50 values,
when given, should also be recorded.  Sources for such  information include
toxicological literature, the Office of Pesticide Programs, EPA files, and
pesticide manufacturers.  A reference list is included  in the Appendix B
and may be useful in obtaining these data.

     If all values 24, 48, and 96-hr LC50 for the specific pesticide and
aquatic species are available and there are no conflicting values, use the
data as it is.  Where conflicting data occur and can be traced to the con-
ditions under which the tests were performed, use the results most similar
to the field conditions being studied.  For example, results may be given
for temperatures of 7  and 20°C.  The temperature most  closely approximat-
ing an actual average stream temperature should be used.

     (Optional) Standardize values from static bioassays of calculated con-
centrations to flow-through bioassays of measured concentrations by the
following factors:

     Static value x 0.71 approximates Flow-through values

     Calculated concentrations x 0.77 approximates Measured values.

     Frequently, data lists do not provide all necessary LC50 points.  Use
one of the following methods to extrapolate missing values in the data.

     1.   Find the slope between concentrations at the  24, 48, and 96-hr
          LC50 intervals for a genus or family related  aquatic species
          subjected to the same pesticide to approximate the missing point
          or points.  See Appendix B for a partial phylogenetic chart.

          Example:(from dieldrin)  24-hr  48-hr  96-hr  LC50

          Given:  Lepomis macrochirus    5.5    3.4     2.8   ppb
                  Lepomis gibbosus        -      -      6.7   ppb

                       Lepomis macrochirus         Lepomis gibbosus



                               =    - = -°125               = "°125
                                                        C48 = 7.3


                     ^ A _ R R   -71              '^ ~ Co*
                      48 - 24  = -VT - -0875     48 - 24  = -0875

                                                        Co, = 9.4
                                     123
-------
                                      = -0375      rr = -0375

                                                        C24 = 9.4

          Now concentration values for 24 and 48 hours have been generated
          and may be used to approximate experimental  numbers.

     2.   Use the geometric mean of the toxicity slopes for other species
          of the same order exposed to the same pesticide.  Calculate the
          missing values using these geometric mean values for 24, 48, and
          96 hours.

     3.   Use the standardization factors suggested in the Federal Regis-
          ter (1978) to estimate 96-hr LC50 values.  The reciprocals of
          these number can be used to derive data points for 24,, 48, and
          72 hours.

          Given:  24-hr LC50:  multiply by 0.66 to estimate 96-hr LC50
                  48-hr LC50:  multiply by 0.81 to estimate 96-hr LC50
                  72-hr LC50:  multiply by 0.92 to estimate 96-hr LC50
                  96-hr LC50:  multiply by 1.52 to estimate 24-hr LC50
                  96-hr LC50:  multiply by 1.23 to estimate 48-hr LC50
                  96-hr LC50:  multiply by 1.09 to estimate 72-hr LC50

     If the pesticide of interest has no toxicity data for the species,
several alternatives remain to find data indicative of the probable toxic-
ity trend.  However, without experimental results, the accuracy of these
estimates cannot be assumed.

     1.   Find the toxicity range of genus related species and use this
          range as the most likely to include the actual values.  A geo-
          metric mean may be calculated, if desired, for computer handling.

     2.   For no genus related species, use family, then order, then class
          if necessary to obtain enough data to describe a toxicity range.

     3.   Use toxicity data from a related pesticide for the same species
          or at least the same genus to approximate toxicity if the two
          pesticides are known to cause similar biological effects.

Step 2:  Sublethal Toxicity

     Find the MATC (maximum acceptable toxicant concentration) range for
the pesticide and aquatic species in question.  If the range is specifi-
cally given in the literature, use it.  Where chronic data is given and
includes no effect levels, use the highest no effect  level and the lowest
chronic effects concentration to determine the MATC range.

          Example:  Highest concentration with no effect - 50 ppb
                    Lowest concentration with effects - 100 ppb
                                     124
-------
                         50 < MATC < 100 ppb

     A MATC range may be found for another species for the same pesticide.
The next two steps describe the procedure to use this MATC value for deriv-
ing one for the species in question.

     1.   Calculate an application factor (AF) from the known MATC of  the
          other species.

               MATC (other species)/96-hr LC50 (other species)
               = Application Factor Range

     2.   Use the AF to calculate the new MATC.

               AF range (from above) x 96-hr LC50 (of desired species)
               = MATC range (for desired species)

     Alternatives may be used when no MATC ranges or chronic data can  be
obtained.  These include the following:

     1.   Use an arbitrary AF based on the pesticide's known persistence
          and cumulative effects.

          a.   If the pesticide is neither persistent nor cumulative in
               toxicity, use 0.05 as the AF for a 24 hr average or 0.1 for
               a temporary concentration level.

          b.   If the chemical is known to have persistent and/or cumula-
               tive toxicity properties, use 0.01 for the 24 hr average or
               0.05' for a temporary AF.

          2.   Where a MATC range is given for the species for a related
               pesticide, the use of its derived AF may be a better esti-
               mate than an arbitrary AF.

     The MATC range may be converted to a single concentration value for
convenient computer manipulation.  Either endpoint or a geometric mean are
possible choices.  The lowest value in the range (the highest no effect
level) is preferred as the most conservative choice.

Step 3.  Input to the FRANCO Model

     The choice of curve formation depends on its intended use.  FRANCO
will determine the number of times a curve is exceeded, the duration of
each exceedance event, the total time of exceedance, and the percentage of
total time that the curve is exceeded as the fraction over SERATRA cutoff,
i.e., total event time.  Suggested curves are included below for each
aquatic species being studied.

     1.   A SERATRA cutoff.  This one is discussed in Chapter 4.
                                     125
-------
     2.   The MATC as one of the endpoints or the geometric mean.  Use of
          both endpoints requires two curves (see Figure 13a).

     3.   The LC50-MATC conservative curve indicating the greatest chance
          for lethality (see Figure 13b).

     4.   The lethal curve least likely to indicate lethality  (see
          Figure 13c).

     5.   The intermediate curve (see Figure 13d).

     6.   The MATC line from 96 hours to the end of the test period.

     7.   The LC50 or LC50-MATC line beyond 48 or 96 hours.

     FRANCO accepts a maximum of six duration-concentration pairs of points
to define each curve (function).  Time is in hours and concentrations  are
expressed as Kg/m3 (1 x 103 mg/1).  Pair point choices are dictated by
the selection of each curve.  Each line segment must be defined  by two
points.  For straight line sections such as the MATC, designation of two
separate durations at the same concentration is required.

Step 4.  Interpretation of Results

     The Exceedance Summary produced by FRANCO lists the following values
for each curve:

     1.   The number of times curve concentration levels were  exceeded.

     2.   The total number of time steps.

     3.   The total percent of modeling time in which the curve  was
          exceeded.

     Additional information obtained by FRANCO is described  in Section 4.2.
Particular points of interest may include the highest concentration  level
reached, the time steps in which an event occurred, and the  percentage of
time steps lasting longer than some time dx at Concentration Cx.

     Results may be briefly summarized as illustrated in Figure  A.I.
Section A, below the MATC line, is presumed safe  and is calculated as  the
percentage exceeding the MATC (sample curve 2 in  Figure 13a) subtracted
from 100$.  Potential lethal and sublethal Section B, which  includes
lethality less than 50%, is calculated by subtracting the global  exceed-
ance determined from sample curve 4 (Figure 13c)  from the MATC curve
(sample curve 2).  Section C measures lethality as exceedance  of the LC50
curve.  Sample curve 4  is used in this example.   Sample curve  3  (Fig-
ure 13b) may be used instead if desired.
                                     126
-------
           ppm
            12
                         LC50
            10
                                                    MATC
                     I      I      I
                    24    48    72    96    120   144   168

                                     Hours

               Figure A.I.  Summary of global exceedance results,
                                   REFERENCES

Baca, R. G., W. W. Waddel, C. R. Cole, A. Brandstetter and D. B. Cearlock.
1973.  EXPLORE-I:  A River Basin Water Quality Model.  Submitted to U.S.
Environmenta1 Protection Agency.Battelle, Pacific Northwest Laboratories,
Rich land, WA  99352

Oonigian Jr., A. S., and H. H. Davis, Jr.  1977.  Agricultural Runoff Manage-
ment (ARM) Model User's Manual:  Versions I and II.  Submitted to U.S. Envi-
ronmental Protection Agency, Environmental Research Laboratory, Athens GA,
Hydrcomp, Inc., Palo Alto, CA.
Olsen, A. R., and S. E. Wise.  1979.
Concentrations for Risk Assessment.
 Frequency Analysis of Pesticide
Submitted to U.S. Environmental Pr o -
tection Agency.Battelle, Pacific Northwest Laboratories, Richland, WA.
Onishi, Y.  1979.  1979 User's Manual for EXPLORE-I:
Quality Model (Hydraulic Module Only
Protection Agency.Battelle, Pacific Northwest Laboratories, Richland, WA.
  	A River Basin Water
   Submitted to U.S. Environmental
                                    127
-------
Onishi, Y. and S. E. Wise.  1979a.  Mathematical Model, SERATRA, for Sediment
and Contaminant Transport in Rivers and its Application to Pesticide Transport
in Four Mile and Wolf Creeks in lowaTSubmitted to U.S. Environmental
Protection Agency.Battelle, Pacific Northwest Laboratories, Richland, WA.

Onishi, Y. and S. E. Wise.  1979b.  User's Manual for the Instream Sediment-
Contaminant Transport Model, SERATRA.  Submitted to U.S. Environmental
Protection Agency, Battelle, Pacific Northwest Laboratories, Richland, WA.
                                    128
-------
                                 APPENDIX B

                   TOXICOL06ICAL PROPERTIES OF PESTICIDES

     The following tables are divided into two sections.  The first  is  a
compilation of toxicity data on individual pesticides and the second  is a
listing of groups of related pesticides along with properties of persis-
tence, solubility, and Kd values.

     Table B.I lists pesticides by their common generic names and func-
tions.  Several trade names are included to help identify the pesticide,
but the use of such trade names is not intended as an endorsement of  the
product.  The table has been divided into the taxonomic classes of fish,
crustaceans, and insects.  While the risk assessment procedure of the CMRA
Methodology is not sufficiently sophisticated to assess the toxicological
risk to invertebrates, they have been included for toxicity data compari-
son.  A phylogenetic chart shown in Figure B.I will help identify the
relationships between common fish species and may be referred to during
the assessment process.  The list in Table B.2 identifies the scientific
names of common fish.

     The individual pesticide tables include information on lethality as
24-, 48-, and 96-hr LC50 concentrations, MATC ranges, and other acute or
chronic effects.  Occasionally, experimental conditions will be noted.
All values are given in mg/1 (ppm), and no attempt has been made to  judge
them for accuracy or standardize them to flow-though bioassays of measured
concentration.

     The reference bibliography is located at the end of this appendix.
Not all citations are from the original articles.  Further search for the
original article is recommended for a better perception of the bioassay
techniques and water conditions used.

     Table B.3 divides many pesticides into chemical classes and lists
solubility, Kd, soil and water persistence for them.  Soil persistence  is
frequently listed as a half-life value (T/2).  Water persistence values
listed as percentages should be interpreted as percent remaining effective
after the given length of time.  The chemical family names used here
include organochlorines, organophosphates, carbamates, phenylamides,
phenoxyalkanoates, triazines, and a miscellaneous class.  While chemically
related compounds usually exhibit similar properties, individual members
may display wide variation.  Therefore, the generalizations stated below
may not be applicable to all pesticides in the group.
                                     129
-------
Organochlorlnes

     Chlorinated hydrocarbons, cyclodienes, and other oxygenated chlorine
compounds are grouped in this category.  These insecticides are persistent
and most are rapidly assimilated by organisms in contact with them.  Most
are only very slightly soluble in water; typically less than 1 ppm will
dissolve.  Their high affinity for and solubility in lipids is the basis
for their bioconcentration in the fats of an organism.  Because of their
low water solubility, they are highly immobile in soils where they are
held and inactivated mostly by the soil organic component.

     Most of the organochlorines are very persistent with half-lives of
several years.  Their degradation in soils takes place at a slow rate but
can occur by microbial metabolism, photodecomposition, and by chemical
reaction.  Degradation products of some of these are also insecticidal,
prolonging their effect in the soil.  Volatilization may be a signficant
method of removal from both soil and water.  These pesticides should leach
only a very short distance through the soil profile.

     Entrance into the aquatic system from agricultural lands occurs
almost entirely through soil erosion and watershed runoff.  Once in the
stream most will settle to the bottom and be associated with the sediment
where it will equilibrate with the small dissolved fraction.  Persistence
in static water bodies is very long and toxaphene has been shown to be
active after ten years.

     Organochlorines as a group are highly toxic to aquatic animals and
manifest an acute median lethal concentration less than 0.5 ppm for almost
all of them.  Biomagnification by uptake through the food chain or direct
adsorption from water is not  likely to occur in moving river communities
(Macek, 1977).  However, where biomagnification and transfer does occur,
resistances to a pesticide by a species may have drastic effects on the
next trophic level.

Organophosphates

     Organophosphates have demonstrated toxic effects to a variety of
organisms and have been used  not only as insecticides, but also as herbi-
cides and fungicides.  Their  advantage over organochlorines lies  in their
transience in the soil with residues disappearing between crop seasons.

     The properties  of Organophosphates often cover a wide range.  As  a
group they are slightly soluble in water but can vary from less than 1 ppm
(Dioxathion) to 13%  (Trichlorfon) or more.  Movement  in soils  is related
to their solubilities with the most mobile usually being the most  soluble.
However, they tend to adsorb  strongly to soils and are seldom  leached
below the surface.
                                     130
-------
     Vapor pressures for this group exceed the organochlorines  and  they
are readily susceptible to microbial and chemical degradation.  The rate
of degradation rises with increases in soil moisture, temperature,  and
acidity; with hydrolysis; oxidation; and volatility contributing  to the
low stability.  Under moderate conditions organophosphates remain in  soil
no more than several months though parathion has been detected  in sandy
loam for up to 16 years (Guenzi, 1974).

     Most aquatic contamination will result from runoff entering  the
watershed.  Organophosphates tend to hydrolyze readily and have a compara-
tively short half-life in water.  Prior to their degradation they are
highly toxic to aquatic animals, several compounds are responsible  for
lethal effects to crustaceans around the 0.001 ppm level.

Carbamates

     Carbamate compounds have been divided into the methyl carbamates, a
thio- and dithio- group and the carbanilates and are effective  as insecti-
cides, fungicides, and herbicides.  Insecticides act by contact or  inges-
tion as competitive inhibitors of cholinesterase.  Fungicides are useful
in controlling foliar disease in agricultural crops.  Herbicides  are
effective in preemergence application.  Most methyl carbamates  have insec-
ticidal activity and most phenyl carbamates are herbicides.  Thio-  and
dithiocarbamates are used as herbicides and fungicides.  Carbamates have a
relatively short residual life in soils and are readily degraded  by most
nontarget organism.

     Carbamate compounds are somewhat volatile and most are broken  down
rather quickly by microbial and chemical degradation.  They do  not  sorb
strongly to soils and are easily leached.  The carbanilates, however, are
relatively immobile.  Once in the water, they are quickly broken  down with
a significant reduction within one week.  Most are moderately toxic to
fish and very toxic to crustaceans.

Phenylamides

     Amines and an-i lines, the nitroanilines, and the ureas comprise this
group.  Alternative names for these chemicals are the acetamides,
acylanilides, toludines, and phenylureas, among others.  Nearly all  of
these have herbicidal action and are metabolized similarly.

     These pesticides show moderate soil persistence with significant
microbial degradation.  None are very soluble, and the nitroanilines  are
particularly insoluble (1 ppm or less in water).  Several are somewhat
volatile and their adsorption is directly related to the amount of  organic
matter.  The nitroanilines are fairly immobile in the soils but the ureas,
amides, and anilines do not sorb strongly and desorb rapidly.  Toxicity of
this category to aquatic animals is low.
                                     131
-------
Phenoxyalkanoates

     This group of herbicides also exhibits low toxicity to aquatic  ani-
mals.  The phenoxy compounds are not readily adsorbed to clay minerals but
will sorb in limited amounts to organic matter.  They are slightly soluble
and are easily leached.  Their persistence in soils depends on soil  water
content usually ranging from one to six months.

Triazines

     Triazines and triazoles are largely nonvolatile herbicides.  Their
wide range in persistence from two weeks to more than a year depends on
soil type, soil moisture, and application amount.  Adsorption, which is
readily reversible, depends on soil composition, moisture, pH, and temper-
ature.  Some are moderately mobile while others demonstrate little leach-
ing or lateral movement.  Toxicity of these pesticides is low for fish and
may be moderate to invertebrates.

Miscellaneous Pesticides

     Included among the miscellaneous are such groups as the aliphatic and
benzoic acids, the dipyridyls, nitrophenols, phthalimides, pyridines,  and
the uracils.
                                     132
-------
o
I—.
en
CTi O
,— 1 • — C
cr» HI
«— ) 4->
QJ QJ
QJ ^£ •>-
C ••- ••-
QJ -O .Q
•r- 3 QJ
Q. CQ 3:
VO
x> cn
•— 2

i- QJ
4-* QJ
QJ 4-» .*

aj .0 -0
Z < CO
f-H
,_!
re
4-*
QJ
_^
QJ
S
en
en


o o

ID ID
i/i t/i i—
o o «



o

ID
yi
o
x:
en
r-~
en
o
o
ID
c
^

<1J
•a
•-
,_
-Q
QJ
en en
in in
QJ QJ

toS Io o o
en cr» en i/i co
QJ QJ
N fsi hJ T3 "D
«2 *£ *
en
in
•O
QJ

0^ vo
co en r-*
QJ ,-1
TJ <*J
S58
en
OJ

c
o
co
QJ
~O
i
QJ QJ QJ
Q. O. Q.
(_) i_> O r^ ^JD r»»

o« »a «»-H ^-t-H
CO (/I I/I (/I CO CO
QJ QJ QJ QJ QJ QJ
"O "U "O "O "O "D
OO OO OO LO i/l oO
a
in
en


tu

,_
4-
CD
C C


00 00

QJ QJ
CO Ul
QJ QJ
•-3 n
        a>      •-<
 CD

 E
                                                                   ^D i^» m o
                                                                   CO CO O 00
                                                                   **• o oo o
                                                                   o o o  •
                                                                                                          CM CM

                                                                                                          O CM

                                                                                                          O O
X

O
oo
^
<:
UJ
o
<

oo

a:
o
                       S

                       CD
                       V



                       O

                       d
                     o

                     o
                     CD


                     CD
                                   r ro      r-H
                                            in P"-^         co in CM


                                         OOOOOOOOO

                                         ooooooooo
                                                                                                      5 ro O O      CT* v£>
                                                                                                      > ^H oo O      ^-4 m
                                                                                                      > O «—" oo      O O
o


o
a
I—I

C_3
t—I


1/5
UJ
CL.
CQ
                                                                        Scr^
                                                                        ro
                                                                     Sf—( CM
                                                                     o o

                                                                   odd
                                                                                O* CO ro
                                                                                CM CM C\J
                                                                                o o o
^H O

do
t-t O

o o
QJ
•a 4->
i- 4->
O O
E E
oo re re
•— • 00 00
re
o. u
U V>
QJ
CO i —
C -C
•- a.
S- (U
O &_ Q
•ochirus
u
re
E
S


en
LO re
£!
cio
S-
OJ
-a
L.
re
o
e
l/l *O
i— i OO

CO
3
O
Dromelas
•ochirus
u
_C i-
Q. E
QJ O
•r- QJ
Q 	 1
-C >>
° 5
4-1 » 00
^ fc. oo c c:
CC O — • O O
t-
QJ
-o
O
E
re
CO
3
4->
re
i_
re
re
re
jromelas
.iculata
> aculeat
QJ 3
l/> l_ QJ
QJ 4->
CL--- QJ
QJ U 4J
clc?o
3
'QJ
&
i
ol
•ochirus
1
§
QJ

oo re 3
5Ea
ell
= 22
> serrula
ncaudus
71 OJ
'io JD
^
13 -3
8-
- «
ustris
kadiake
(O dl
^ d)
^1
JcZ
grandis
QJ
OO QJ
OO Q.
C
0
*4-
u
&
c
o
I-
-^
)acif ica
z
Q)
C
o
«
en
•ochirus
•o u
•"- E
O E
x e I
oo re QJ
CO
                                         o o
                                                                o; wi
                                                                o e
                                                                         133
-------
                                                                                                                                             o o o o o
                                                                                                                                                                 O O O  O  O O
                   V)  (/>
                        i-  i-
                        QJ  QJ
                        ~   "
                                                  yl 4
                                                      _
                                                      QJ
    OJ  co  in  v> in vi


                                                 QJ Q)  QJ  QJ QJ Q)


                                                 C C  C  C C C


                                                OO CO CO OO OO OO
                                                                                                                                   j^        -^r^^Jurjiirj^E-^J^Ji^J^-ii;-^
                                                                                                                                 MQJNMQJQJQJQ)QJ>'-(U(IJ CM O  OOO

                                                                                 oooooo^roo    ro ro o  o o o
CQ
                                                               O CM O Q O O
                                                               ro ro o O o O
                                     §
       V> ••-  VI        i
    X    U  3        «
    QJ V> V)  U
   i— 3 re          i

CJ •»- QJ 1_  i_    CO i
<  c o re  re    t— i
i— -c o E  £    o i
co  Q. E E  E    LJ i

o: o co o CD    z c
LJ                 •-«
                                                     .
                                                   O- *
                                                    -
  re > •«-  (j     re
  c:     >  vi  v) c
  01 v>  QJ  re  QJ
)  re ••-  i_ 4-  4-> m
r  E <" -O     o> QJ
:     a.     vi  c •*-"

5  * -8  =  £  8 ^
:  c ••- i—  re  QJ c
•  .c t. *"•  £  re O
>  Q. p.  QJ  E  •— O
j  ro >^  vi  re  re i-
:  o c_>>-^
                                                                                                                                                                               ^
^-  C O  O  V) (J
 QJ  3 1-  t.     (/>
 E  CL O  u  v» QJ



 3  3         Q>»—

 3  3 '«-••-  O.

•— *~ i  g  p re
 re  re O  o  i- u
4J 4-» Q.  D.  U L.
 U  U QJ  QJ  •"- QJ
                                               •sz
                                                                                                                             CO -r- -t-
                                                                                                                             O *-»  U
                                                                                                                             i  u ••-
                                                                                                                             a.  QJ  i-
                                                                                          134
-------
                                                                                                        o    o o o o       o o
                                                                                                                                                 o    ooo
                                                                                                         QJ     QJ  QJ  QJ QJ
r-H i-H
r- r^. CM CM CM
01 Oi r*-. i-- r"-.
'£'"
C C QJ QJ QJ
dJ OJ TJ T? T3
e E c c c
•^ -r- re re re
O_ O- l/l t/l l/l
CM CM
r-. r^.
Ol 01
CM
,-( QJ QJ
•o -v
CJ C C
o* *o re
3 00 I/O
QJ ••- ••-
O-H- <4-
o 3 3
o re re
to CD
CM CM 00 CM CM 00 CM
.-(.-I QJ QJ i-H t"H QJi-H
"O V) I/)
(-J C_) C C O <-J C O
O'O"      V)  vi Vt
                                                                                                                             O      r-* r-»  O <-t    .-H
                              K    8^
                             14-     . O
                             M-    CD  I
                              QJ     I  O
                                   O Lf)
                              O    LO O
                                                       CO Ol LD CM «
                                                       §§?
                O O    O O O O O
                 II      I   I   I   I  I
                o o    o o o o o
                ID ID    ID LO LD UO ID
                CJ O    O <_> O C_J O
3 o >) re
3  . re TD
  o T:
>      o
1   LO CM
                                              00    00000
                                                                                                                          4-1 OJ

                                                                                                                          E >4-
                                                                                                                             §«*-

                                                                                                                             QJ
                                                                                       co en ^r
                                                                                       .—i ro ^r
                                                                                       ooo
                          8
                   CM •—t  ^ CM
                   «-H O  O ^H
                   o o  o o
^J- LD CM
ro ir> f^
ooo
 O
 o
                                                                           88
                                                                                                     Ol lO ID «



                                                                                                     O CD t-H *
                                                                                 ) Ol CM VD O       CD




                                                                                     f-H i—t       CM I-H
                                                       O f^

                                                       CM ID
CQ
                           .
                        re *j o
                       ••- »rt -O
                  X ~r~ U
               C   t/1 O OJ
                          «O W
              .  .
           t/i  Q. D. ai E E E
           m  re  re   'r— •


V> J_



»— JT
QJ CL


E 01
O) -i-
£ -C <
CX CX 4-












>,




1
i





re o

i- «*-

u u

•— CL




rs 3
r QJ QJ
3 C C
-00
U J_ S-






o


QJ
JD

QJ



V)
CL
O










QJ
C
•o


re


3. J





re





ex

V)
cu

re

CX
1


VI


re re
c •—


c o
0 1-
<4- CX

v> v>
3 QJ
C • —
•r- re
•— ^
QJ CX
> QJ
i^uo a!








u
o







I
CL



.c



VI






u
c

_c
t-









I-
QJ
C
•o
i.

re
en

1




(/i
re
i/i i —



i- i-
re 3 CL
+j re
4-> t/)
3 V> QJ
1- 3 «—
4-» -^ re
VI -C
Q vi CL





v>


v> re
re 4->
*~ c



Vt V>
3 3


re re

                                                                                                                                                          •g
                                                                                                                                                           i»
                                                                           •s
                                                        •g
           £§
           o_  u
                                                                                                     -g
                                                                                                     SC (^
                                                                                                     ^-. ^-
                                                                               135
-------
CO
en
1-1 OJ
a.
0.0
o
(_J -O
,*g

VI t/1
QJ QJ
•D -0
C C
1
CL
O OJ CTl CM CM
cj r-* vo r-v r-*
en en en en

 ui (/i (/i (/i
OJ QJ QJ QJ QJ
•O -O -O -O -0

JD
in

en r-.
1—1
QJ
0.0
to  r-*
QJ f-H
•o


8
O

o
m

JD
•— t •— < m •— «

en en en en
rsi N cu KJ


oj r-s
oo m
0 0


i i
o o
en
OJ

^~
2
•o


CM
0


0
in
a-aa
ro ro ro
4-> H) 4-J •*->
QJ QJ QJ

»— tn ' — i —
T5 T5 T3 T3


OJ
ro
0

O
1
r-t en
"QJ 4J
QJ
c

VI •—
II






2
ro
•4-*
OJ 4->
JD QJ
c «— i in

vi en en *—
QJ *
T3 N QJ -5


0
g


1
0
2£
o
4-* CJ

.— (/I
II






«— I
Q. re ro
O 4-> 4->

(XI i — »—
QJ X X
IT'S


'o


QJ
S_
r-l t— <
CM
4-> 4-1 r--
«— i

< — r— v)
* * fc
•O -CI TJ


O


OJ
t-

cn CM
f— t *— i

v) v>
1 1
in


d

s
cn
re
4-»

r~~
-1


o


£
                                      §
vo r-- r^ ^j- rv.
in m O ^ *fr
o o o o o
                       CM o o o o en
                       in o en &* ^- in
                       o tn t-< o o o
i*-. O «> •
:§§si
                   o o o o o
                                                   O O CD O O
                                                                          o o o o o o
               Si
CO U^ i—I <* O '
o o o o.-..

d o o o o i
                                                                                .-H O •-»

                                                                                odd
                       O
                       s
o c^ CM ID in i- in
o LO CM in CM -c o
,-H O O O ,-H   O

d d CD o o ^ o
yi -i-
3 to
4J C
ro co
3 3 (/» vi ro

1- =3 4-> i. -O

(/I (J -r- ISt ^ •(-
x ••- u 3 ro
OJ 01 > (/> U C/i C
r- 3 QJ ro ro QJ
3 ' — i- 4- ' — 4-> t/1
o. ro JD QJ cu
jr vi v> c -M
ro o. »/» 3 a o u
•r- QJ 3 t_ J- E 0>
CO' — *O TJ OJ C
Q- E=  re s- »— •
1- JD 0 3
O CJ JD
•u re >i ro
i — i — 1/1
t/i i —
>> QJ ro re
O O *<•" >r~
I- i- i- c:
ro ro 3 QJ
c c QJ (/>
O o c wi
i- i_ o ro
QJ QJ i. ro
•*-> 4-» U r—
D_ D- CC CJ)


O
4->




QJ
O
>>
to
a.
|
i*
JZ
u
U)
4->
J= >1
U 7

3 J=
V> (J
•<- V>
J^ 4->


JJ
O <->
c c
JC -C
s. s-
o o
SII





t-
OJ
c
u

re
en
g
5
to

c;

4->
C
o
^-
w

£
QJ
>
ro
oo
to
ro
i/i i —
3 QJ
+J o E
ro •!- o
1- O. S-
3 I- Q.
vi ro ro
3 O 
"~ to QJ
r^'vi c jl
(/I •!- ' O,
x ro L. QJ
Sfe5,-S
UJ O C-) Q-
 re -— i_
ro •— 3 ••-
4-» 3 0 JZ

c -^ o
3 4-> to l_
°" S! 5 ro
V) 4-> E
3 ro v>
J_ ••- O w
3 •— S- -i-
r- -r- QJ E
re u 4-> 5
4-> QJ (/I CX
o o ro QJ
•-• Q. CD _J
(/>
3
4^
ro
3
i-
1-
QJ
y\
ro x
C QJ VI
cn«— 3
co re 3 ' —
z E o. re
< J=
uj re re a.
o ••-••- QJ
£i.ig
to O. CL E
^3 ro ro -f-
cc CD o to

Vi


3 t,1 ro ra

C 4-' OJ 0
OJ W 4-> VI
u -1 IM ro
4_> cr re »*-
i/i W
ro in
to vt i — 3
O. 11 •— i-
,2^ 41 *°
>, 01 >, re
o «a: i o
in
ui
C
QJ
.^
v) re

s- "o


3
O V»
re QJ
i — 4->
QJ
c/) C
3 O
feE
OJ
g ro
re Ic
CD Q.
                                                                                                              «f
         5
                                                           136
-------
                                  CO CO Cd CO
                                                                   t-H «—« t—I CM
                                                                                                                            oo cr^ co
                                                                                                                            kO Lf> VO
                                                                                                                            CTi CTv CT»
                 ,-H   a)

    C71 C71 C71 C71
    C  C  C C

    J_  L  I- J-
    a>  a>  QJ oj
    .^ .^ .^ ^
    U  (J  O   t/>  l/>
C  3 3 3 -*e
 QJ  QJ QJ
 O- Q. CX

33 3

«a*a *a

 i/>  i/i in
                                                                                                                            en c  crii-
                                                                                                                            c  o  c u
                                                                                                                            •r-  l/l ••- C
                                                                   in ir> m
                                                                   ro t—i CM
                                                                     • CM CD
                        ?
                     IT) O
                     JC. t-
                        .C
                     ^~
                     PO CO
                                  m co co  r-^

                                  CO CO «-*  O
                                                                      000

                                                                      00*0
                                                                                                                      > ^-< co co o t-< o
 o
 o
CO
                                  *r c^ o r-

                                  CO CM CM O
                                                                                                    t-H CO  *-l
                                                                                                    O ^H  O
                                                                                                    o o  o
CQ
<
                                                                                                    000

                                                                                                    O 0*0
                  U  (J 3
                  i-  >> E

                  C  Q- C
                  o  o o
                  i-  J- i-
                                         -
                                  VI r-  03  (-
    L !.•>-  0
    3 Q.4->  1-


    in QJ     E    Z

   •i- re •(-  i/1    LU

    (/i CL-I-  E    
i/i  « ••-  o  0)    3
i-i o o_ a. —I    a: c
u_                  o
                                                      o.    i. •—


•r
QJ
C
-a
u
ra


i

^2o"i
QJ
O
PL i/l
U -i-
4-
W C
g*
O «3


E 3

Z(5
3

^!^
QJ O
C i-
fO t_)
&i



§1
O- CL
CU (U
in

«O
U
V)
•0
CO **-


t.
3
U
re

3
S-


fO
.^
•u
^3
"3

QJ
(J
1-
re

I- 0
O I-
LU QJ
CO 4->
O
»*-

U

&
s_
re

o
S-
QJ
4-1
to
O


VI
re
c
OJ
in
in
to
re
' —
u

i/i

in
O
c
>>

i_
O
3: o
t/l C


.-
OJ
C


en

|

re
U1
>D

4^ ^
i- J-
3 Q-
(/> QJ
3 • —
•i-  QJ 3

• — 3 ••-
3 (J -C
•r- O

QJ 3 U
i_ QJ re
re m
•t~ O l/l

^ QJ E
(J 4-J O

0 * S"



re
C
en
z E


CJ •!-

CO D.
CXL O
tn ^

4J i.
U 3
in o

1/1 1/1
3 3
i- i_


CD CD


.
. 5t~
a.
l/l QJ

c in
CO QJ d
1— en o
LU X TJ
£££
             o    n:  T3 ••-
             cc    i-  ••- j=
             t—    O  U 4J
             •-• _i a:  "~ o
             z >- o  4-» i-
             O Z Z  0  O ©
                                                                                   o ••- •—
                                                                                   X 4-> *O
                                                                                            137
-------
                                                              i—I ID
                                                                 (-*-
                                                               C Cn
                                  .. 10       •—
                                  !•> Oi       w i
                                                  r-«*          di f— co co
                                                                                     oooo       00000000000000000^0*0*
                         _
                    -o-o-
                     33
                                                                                                                                                      coi
                                                                                                              E
                                                                                                              (O

                                                                                                             ">,
                                             U-l LU CD UJ UJ    Li.i '''
                                             UJ O CD U-l O    UJ LU
              CO LU
              CD LU
              a. co
LU     QJ ^ L
CD LU  E •>- t.
CD LU -r-  (J <
CL CO O «£ (
                                            ECQLU  E CQ LU  E CO LU E (33 LU £ -^ QO LU UJ
                                          J't-CDUJ-i-CDLU-'-CDUJ-'-CJUJ'.- (J (-D LU O
                                                                                                                                                 aj "O L
                                                                                                                                                 £ -^
                                                                                                                                                 -
                                                                                                                                                            I «->       CO
                                 00

                                 O O
 O
 O
CO
                                                                                      «-«ir> o
                                                                                      o LT> ^r i
CQ
«=£
                                                                             O
                                                                             d
                  !§'
                  i o.  i
                  - (/I  !
                  1,2,
-21
 O. t-
 i.  O.
             X •— 4J Q. (J
••- 3 3 JZ
C O VI U
•i- •(- O O


*° *" "5»i
         I	 ••-•!-

         o o o
         QJ O. CX
         O   o
             !—(/)•'-
             O  C CJ
                                                                                  138
-------
O CD OJ




       Oi
                                                                                             00    OO
                                                                                                                        00 O O
                                                                                                                        r^. i^, r^ r-.
              cn
       O  O  C     _
O     O  O -r-    -r-
<£>     JD JD  S-     I_
Cn Cn XJ X)  OJ     OJ
•—I tO C  C -^    _i^
   cn re  TO  u     u

•-^           Cu -Ct O-
 «3  i_ OJ  <1)     IT)
    Q) X) X) e3 tO 08
40 4-J -t- -r-     cn
 QJ  (/I (J  (J  1- i—l  i-
    (O •!— •!—  QJ     QJ

    §«J 1-  i-  i- CL  1-
   f—  vO vo
                                                                                                                                                                              CT> CTv CT»
                                                                                                                                                        OJ  OJ  O)
                                                                                                                                                        a.  a. a.
                                                                                                                                                        o  o  o
                                                                                                                  o 10
                                                                                                               C  in O> e
                                                                                                               O  I- ^H
                                                                                                                                                                      .
                                                                                                                                             QJdtQJQjQJQJQJQJ

                                                                                                                                             cccccccc

                                                                                                                                            l/l l/l t/> l/l (/I tO l/l tO
                          8
                       t-H O
                        I O
                       O «-H
                                                                              - •—^        r-t  \        vD
                                                                              >i-ioo    OJ^H    Or-t    ^r^-i
                                                                              JOOO    OO    OO    OO
c
o
u
                                                                                                         3O    OOOOOO
                                                                         000*00    oo    oo    oooooo
                                                                                                                                   SO *
                                                                                                                                   o <
                                                                                                                               o o <
                                                                                                                                                      o o o o o
                                                                                                                                                              §88
                                                                                                                                                              O O CD
                                                                                                                                                 •) ir>        ^H
                o o
                ro co
                                              o o o
                                              V£) •-«<£>
                                                         o
                                                         o
                                                                                                                                                                                  Scr> ^o
                                                                                                                                                                                  r* O
                                                                                                                                                                              o  o o
                                                                         g     o
                                                                         o     o
                                                                                                              8
                                                                                                                                                       OiJ  r-~> 1^
                                                                                                                                                       *—<  *9" ^H
                                                                                                                                                       o  o o

                                                                                                                                                       o  o o
             Z  (j •—
             1/1  c «s
             •-« o to
c

T) X
C OJ
« S1 3
5 E Q.
^
0 -2 .2
i/> 'o.'ci i
a: Q Q o-
fl 10
3 CJ
c
o
^
(O
o

in
o
t.
fD
(/I c
t£
= ui  o
c to <*-


CT>4-> -^
111
to to to

<_) O-


«o

in
3
3

K-.
in
in 3
3 4->
3 4-J
4_> tO i — !_-£:••- i_X)3
•Oin-— 3---CXQV1 l-3*J
c ••-••- o o « u  l/l«3' — J-*t~
m^^4JEEu^ zEO-rajD^

t-'t-'f-Omm-P'*- uj«QfOCLtn3
3 in i — 1- ••- ••- Q- O •*- -i- c
o «n
«*- o

u to
in
in
i_ c

O in
0> TO
a. o
                                                                                              139
-------
                C\J CM CO CM CM
                CT CT> C  CJi Ol
                c  c  o  c  c

                i_  i-  S-  i-  t-

                ^£ ^ -C ^ ^£
                O  (J  C  O  U
                •r- -i-  0) -i- T-
                Q. D. I Q. Q-
                                                                                                                       o o
                                                                                                                       o 5
.   .
 So i— *~<
 o QJ
                    CX £X
                    o o en
                    • > o 10
   v> v) vi r--
   i_ t. i- en
   QJ Q) QJ i—4

C C C C O
                                                                                                                •-* c c c
                                                                   _
                                                                              8
                                                                              o
                                                                                                                                   o o o o a* o o
                                                                                   (/I  (/)  (/I  «/>
                                                                  c  c  c c c c  c
                                                                  rtJ  «J  <13 rt) *O *O  «J
                                                                 l/l (/I CO C/> 1^1 CO t/1
                                                                                                                  •«•£  cx^"§i
                                                                                                                  •-  a  p r:  3
.      -
 5,_ 0)
 <: x
                                                                                       o o    o o o
                                                                                       oo
                                                                                       in in
                                                                                       00
                                                ooo
                                                ID ID un
                                                000
                                                                                        >> :>>
                                                                                        1O to
                                                                                       •O -Q
                                                                                                  so c
                                                                                                  i ro r
                                               O CM
                                               O> CM 00
                                               o o 

3

to
QJ

QJ
E
(U
CX
UJ
c

CX

*•

_g
g-

en
o
1-
o



«o
u

VI
u
1_

c
o
QJ
4->
(X
4->

VI
I-
O
TJ

v>
i-

c
o
4->
Q_
>d


o


to
s-


c
I-
o
c
o
4->
4-1
to

*
s,
v>
o.
e
•o
•>>
x:
u
4-*
3
VI

-*

^
C

.c
t-
SI




fc
c
"£
fO
, en

|





ex

o
VI

c
i-
s
o
VI
3
l-


u
o
s-
^
E


§
&
— 1
                                                                                                                                       re -D  3 ••- •
                                                                                                                                    «J  > •»-  U
                                                                                 1  -r- -Q  3  ;



                                                                                     . p. i V)  <
                                                                                                   s
                                                                                                 ;l

                                                                                               4->  t) VI

                                                                                                3     «
                                                                                                O  V> C
                                                                                                «O  OJ
                                                                                  
-------
    OJ   o o
                                                                        CO
                                                               CT> o^ cr* vo ^
                                                               VT> *O VX3 CT^ *O
"O "O "O  CX "^ ^3
 C  C  C  O C  C
 o  o  o o o  o
CO CO CO    CQ CO
                                                               C C C L C C
                                                               o o o   (/I  t/1  t/1  (/)  V)  
               0.1—
               25
                           CCCCC-— CCCC
                                                                                                            "O TJ "O "D "O T3 "O
                                                                                                             C  C  C  C  C  C  C
                                                                         rn in

                                                                         £££
                                                                         -<-H 01
                                                                                                            t/1 1/1 CO 1/1 UO I
                                                                                                                   &o

                                                                                                                U S
                                                                                                                                                       a. a.
                                                                                                                                                       O O 

£
t/1
CX
••8



0
ifl
TD

>
S
J3




•i
1/1
1
Z3 -i- ••-
•M i- "O
«/) L) f> C
S <£ r^ ^J w

£ 3 3 |^



^5<5c?o





t^)
t-
U
LU
to
zr
ex


•^

J



ro


CX Q.
i/» (/»
> =
O i —
1- ••-
*o ^
C CX
O 0)
I- C

•»-> •(-
Q. _J


a*
,
V.
o

*/} c


C Q.
T3 t/i
I-
••- t/i
<13 3
C7i I_


, — 4_>

                                                                                                             3 > '•- O 3    «0
                                                                                                             :    > t/i u vi c
                                                                                                             n i/) QJ ti TO  (/i
                                                                                                                               1- 3 •"-
                                                                                                                               1- 4-» t-



                                                                                                                            X    U 3
                                                                                                                            OJ t/1 t/1 U
                                                                                                                            i— 3 «o m

                                                                                                                            cxlo *i- •—
                                                                                                                _
                                                                                                              . p. 
-------
                                                                                         CO CO LO LO CO


                                                                                         CT> CTi t—I I-H CTi
                                                                                                              (M CM CM CM C\J CM

                                                                                                              *£> U3 ID UD l£> l£>

                                                                                                              CJi CT^ CTi CTi CTi CJl



JD
•— t .-< t-H LD
LO U3 LD 1O

N Nl N QJ
re as re o
^ i^ :*: e>


So o


ooo
1 1 1
ooo
LD LT> ID
o o o
_l _l — 1

i- 1- 1-
JZ £ -C
CM OJ OO

re

CD

c
o
VI
QJ

QJ
z:
















re

QJ

§
in
QJ

QJ
















o
CD
•JE

_j

cu
_J



(J

c
o

jz
(J
1
1^1


^.
ro





O
P-.
l~l
ID

1
O

0

+j
re

VI
4-J
o
cu

QJ




&

N
'£

O
CO
I— 1
o

0

0
ID
CJ
— 1



CM

£
QJ
O
OO J3
cn
c

(j O


o
o

o

o
LJJ


re
•o

§

i-H t—l T— 1 i— 1
QJ QJ QJ QJ
Q. Q. O- O.


*3oBoao3,-i^.-i.-t,-i
ft)ft)ft)ftJQJQJQJQJCU
•OTTTJTJ'O'U'O'O'O
00 00 00 00 00 00 00 00 0^


O O
ID ro
CM PO

d cj
1 1
0 0
LD LD
O O
UJ UJ

i_ L

*3" O
LD ID
QJ QJ
Q. O-


V) t/1

T3 T3
re re
OO OO
















c:
C

CD


VI
1


CM
O

O
1
o
LD




re

o

CI f- r— .— r- f— F—
4- O-
rt3 <_J QJ QJ QJ QJ QJ T3 -^^^^^Ji^Ji£J^
CD re -i—>(— -r- -i— >^> •>—
"^ in C7t CTi vi vi
CU QJ QJ QJ
-O T3 QJ QJ -O -D
J! 35 JJ
















 o
               CO ^-< 01


               o o o o o o

               o o o o o o
ID LD CM
r-i O O
ooo
                                                                   o o o o o
                                                                                         o o o o o
                                                                            •) CM O  O O


                                                                            i o o  d o
CO


UJ

_l

CO
                                        CT>    ^f
                                        CO    00

                                        s    §
                       ) O
                       •) ^
               ,-4 O .-H «-<
               ooo o

               ooo o*
                                         u  3 _c
                                         re  vi  o
                                            o  o
                                  re     3 ja
                                i —     4).,-
                c re


             oo re re
            l/l  c
            •—  O
&&o,    ~s~


            WJ Q.-I-
             Q_ Q_
                o o
                u u *
                c C «
 (DEE
 -*-> 0 O
 fl Q. O-
 re at en
                                                 I- -O 3 '•- -r~
                                                 I- 3 4-» i. T3
                                                 QJ re re 4-1 re .


                                              x   ••- u 3

                                             ii— 3 QJ re re CD
                                                                  j t— re  re QJ  c
•- (_) V)
3 -^ O

'« «

>, re re
I- C

: QJ c/i
C VI
o re
VI
3
4->
re
i-
re
V)


V)

oo re
m re
re +J
.— re

l- •-
Q.-4-J
QJ
V) 1-
QJ
.— re
re •«-
JZ ^
Q.-'-
QJ U
£8


S
re
u

VI


o.
cu
V)
QJ
V) "O
1 1

u re
(J V>
IS
QJ

•r- Q.

Q. U
QJ -^




OO

UJ
CJ

t—
OO
(/)
3
4->

O
in

**"

3
U
re
1
V)

V)
3
(J
re
1

13
i-
re
re


cu
c

£

en


oo re
I
.c
o

E




Q.
QJ




00


CJ
<
1—
00
QL
re
13
T3
re >
c
ty\ vi
g «
Q.
re o

C -r-
JZ t-
CL p.
re >>
Q 0
                                                                                                                G-S
             a: u
             O QJ
                                                                                       142
-------
8
CO


UJ
                J QJ QJ

                3 T3 -0
                                        . r-    QJ QJ r-» QJ P-.



                                        ^ i—I Cn **- *f »-l -i- t—I
                                             t-H O- Q.    Q.
 QJlOQJQJQJCQJQJQJQJQJ

^"re^3^  E "£"£^ "£ ^

                                                                                               c:  c  c  c
                                                                                           c  c  c  c
                                                                                                                C QJ C •— QJ
                                                                                                                                     t-H .-I t-H       00
                             eJ  U  E -U



                                                                                                                                                              tJ>  O_ CJ 00
                                                                     CL  3  CL-—
                                                                     CL-O  CL-—
                                                                     t—i  re ro  re
                                                                     re  x  re  x


                                                                     CO    CO
                                                                                                                                     •-4 00 CO
                                                                                                                                     c\j in n
                                                                                                                                     o o o
           88S8
                                                                                           ) o O o <
                                                                                           J O O O C

in

in
§
in _i*
D in re
•O 3 ••-
« "re "re m
o ••- _* ••-
••— CJ «O
£in in c
re QJ

.Q QJ QJ
in c 4-»
4/1 3 O U
=9 U E QJ
QJ H *— (J
m re re i-
re
.c
u
in
4->
r
re
^

QJ
in c
3 "0


>, cn
JZ
fei
31 O •—
oo c re
•— i O OO




"C in (/>
•.- re z>
V) O •— t-
5-^ ^!c
re i~ o u
S.  ' —
3 .,- -r- ID 
in 0


QJ QJ
*J 4J
CL 1/1
es
U -I—
••- •*->
Z OO





re
u
QJ

re
re
^
Q)
re
>,
z






CL


in
•*->
QJ
ITJ
oo .0
t— -r-
(_) f—
UJ i —
oo re
Z CJ








CL CL
m i/>
re «3
«P3
•— QJ
re ja
c *•-
L.J 	 |






CL

QJ
1/1 re
3 "O
j= 13
CL CL
QJ •.-
11
-.- 
^ -^ CL
s- E O
0 0 L.
U CL U
C QJ ••-
                                                                                  143
-------
                                   O       I-H
                  in  i/i  wi yi
               X i-  s-  J- i-
               J3 Q)  QJ  QJ QJ

               o c  c  c c

               l_> CO CO CO CO
r-.

I-H

t_
0)
40
1
(X

m

o

o
in
O


JT

o
OJ
**



TO
QJ
QJ
s:



^



J->O
0) O

'D.O

'iy
o
CTt
t— 1

t-
QJ
4->
QJ
O
CO

in
CM
8
O
1
0
in
o
CO

JTO
O
o •
CM O
c
Q)
X

00

«-
O O1
oT""1













o



i_
•«->
s
JC
CO













o cr*
r-. t— <
CTl t— i

r— cn 10
OJ """ ,-H
4-» QJ QJ «/>
§J* C QJ
QJ QJ T3
JT DEC
CO i Q- (/I
1
m

o
o

o

o
in
i_ ro
-C in
0
o •
CM O
o
fTt
«— i

t-
QJ
•M
QJ
O
.C
tX

1
OJ
UD
O
O
1
O
in

	 i

t— f^.
0
o •
OJ O
l_ 1_
QJ QJ QJ
D. •-* ^ ^e
or*. o o

*a ^ c£ S c£ e£ ^ c£ <£ c$ ^
in r- QJ
QJ ^HCQJQJQjaiQJQJQjQJQj
c oe^^^f— i-^^^t.
10 3^0.3333^33300

















*~*
OJ
1
o.





1
o
m


t-

\o




•^
QJ
C
O-



8
A

0
in
y

i-

^>



VD

in
QJ
TD
CO

8
•-4
4-1
X


QJ

£8

*n t-

^•sc
                                                                                8
 C
 o
 O
   80
   t£>
l-v. ^H

o o
                                           8
O tn r^   ro ^j-

   SO o   o o
   O o   o o

o o o   o o

CO
'•*
r-4



O
CO
ooo ro
CJ «~i O



CO O
CM O



in co cr>
CM oj ro
8O O
O o
OOO


,
m *3-
ro ^~
00
O O
00


,
r- 1
o
ir>
o
0
CO
CNJ O
0 O



                                                                                              o
                                                                                              o
                                                                             §
                     tn •>- (/i
                TO X    U 3
                c aj in (/> u
                O»r— 3 TO TO
            CO  TO 3 •— »*- *—
            Z  E O. TO
            5       J= W W
            UJ  TO TO O. 3 3
            (j -t- >r- QJ J- S_
            <  C C U « «J
            t- J= J= O E E
            to  Q. a. E E E

            Ol O O OO CD CD
                                                    OJ   •—    -r-
                                                                                   Q CD    Z K-.
ifornica
chawytsci
in
in
1 ZJ

c
: jz
fe
) *T~ (_)
l/l C
,-H O

o
Q.
(J
I


5
Ml

3
Qu
0
£
O

i
§
Q.
QJ
_J
-------
                                                        vo to to 
 «— i m  c  c    c •— t in c
 lCVO  O  O   v» < — vt O CT* vi
      i  L.  l_ -^ 1_ •— I t
        QJ  (D 1- QJ
       ''

                                                                                            CD      ,-t .-I r-t
                                                         ccccccccc
                8
                                 O CO
                                 in ID
                                 O ro
      CM O

      •-H O
      O O
      CD O
 C
 O
 O
                     CO
                •-H   i-H CO ^O
                  Scvj •-* in co en o <
                  •-I O O O •-! r-l t_ - -  ,
                ooooooooo<
                ooooooooo*

                oooooooooooo
                                                                 LT) cvj ^- ro r^
                                                                    §O CNJ O O
                                                                    O O O O
                                                                 O O O O O

                                                                 O O O O O
                                         *f 00 (£)

                                         O O O
                                         ooo

                                         O O O
                                       O ID
                                       CM CM
                                       oo
  .
^H O
o O
O O
CO
                ro o    r-- co
                ^H CM    o •-«
                O O    O O
                o o    o o
                                    8
                                      8

                       i
                ^ ^  5-  l-
                UOTO-i-
                C C^-  (O
                 j O «— r-

                 I C  re  re
                 3 O OO GO
        *o

        3Q
          VI
        vi re
        3 T—
       ••- re
        w.e
                                    4J 3

                                    S^
                                       ere
                                       •--
                                    3 r~
                                       -
                               > QJ re
                               . E 4-» i

-S
(O
c


0
•^
_£
D-
Q
I- T3
I- 3
VI O
re x ••-


re 3 •— !_
E Q. re JD
•!—•<— QJ 3
.C j= 0^
O. Q. E CU
00 00  VI
4-» S-
•i~ VI
U 3


H- ' —
I- I-

E t
CD C3
3
~% ^
•* "^

 V)
CU QJ
§ QJ
re o
f— o
Q- 0
2
_Q
re


OJ
u
re
t^ c
I- 0
CJ 1-
UJ O)
Z Q_
i-
o
^
re
u

V)
S-
c
o
 j- 1- E

1/1 E E • —
CC CD CD CX
VI
QJ VI
E •>- ••-
O C 1.
i- -i- eg
Q.M- C
**- -D
vi re i_
a; ••-
re •*- en
-C VI
a. 3 c.
QJ -0 6

«— i D_ CD l/l
V)
"re
re 'o
C VI
CTi *O
on re <4-
5 E „
uj re 3
o •-- i-
f£ 1
on Q. E
QC Q CD
                                                     o •*— •"— *f~
                                                                                                             iJI
                                                                                 145
-------
                                                                                                                                   CD O 0
                                                                                                                                   r-.rv.r-.
                                                                                                                                   en CD cr>
                                                                                   QJ QJ
                                                                                   O. CL
                                                                                O O OOOOO^O
                                                                                r--(_>(_)r-r*.r^iOr^
                                                                                                                    V- >-    k-k-    '.- >_ 1_  »-
                                                                                                                    QJ QJ    QJQJ    iU QJ QJ  QJ

                                                                                                                                    i/l CO VI  C
                                                                                                                                   •<- ••- ••-  OJ
                                                                                                                       " V£>
                    -       -    _
              .-< i-H  0) iO «-H  QJ ID QJ
              CJCJ
              O* O*
              3 Dt
 -i—ire
srsrtJSicL   :EZS:O-    a,   z:
•a
-p
 c
 o
 o
i-t   CsJ

o   o
                                                                                                                 rt    ™
                                                                                         8 § 8 2 8
                                                                                                                         O
                                                                                                                         00

                                                                                                                         o
ca
               C71 U rtJ   i. '

               BO-&   5:
               E E OJ   re
             = ^^ 6   t4
Ul
IS

re u
C t/l

•/>
re 3

c «
JC E
D. E
C
QJ

i- -o
4_> re

u i/i
re QJ
"~ 
3
S-
-C
(_)
o

u
E

1 I/I

1
_J
l/l
re

s
re x .
c: QJ t/i
co re 3 •—
z E CL re



CO CL CL g

t/i
m T3


>
.£ S!
t/> JD
CL


0. QJ

C
QJ

•tJ 1- X3
U 3
t/> (J «/>
v "~ tu



II S
CDCDC^
ornica
4-
re
u

u


CO C
I— o
CJ l-
LU QJ
CO 4->
3Z CL.
U
3



_C J-
(J -r-

>, Cn
^
fel
z o •—
co c re
I-H O CO
re

i
o
L.

*->
4-> I/I
3 OJ
1- •—
4-> re

o a.
Ig
re ••-
CO Q.
                                                                                                                                      .— c ••-     o
                                                                                                                                       i g. «S
              •S
                                                                                   -g
            ssli.
                                                                             146
-------
o o o
r-» i-- r-.
O\ CTi CTi
J- S- S-
in
<
£
00
QJ
O
z:
t/i in
<<
££
"QJ"OJ
u o
s: z:
*-O "-O
LO 1C
QJ QJ
CL CL
O O OJ CM Cft OJ OJ
o o r-. r-. 10 r- r--
(/> i/i in wi i/» 10 in
c c c c c c: c
OO CO CO OO OO OO OO
CO 10
S2
c
QJ -f-
cm-
O 3
o re
cs
00
l/l
1- C
QJ QJ
T3 in
C C
OO --3
i— 1 i— < LO t— «
N N  4-> Ct4->

LD LO LO
re re re
QJ QJ QJ
c c: c.-i ro
o o o 10 in
J- 1- i- •-! i-H
QJ QJ QJ
•0 TD T3 M QJ
C C C -M CL
:r


ss tsti
OJ
ex
o
o0
1_
QJ
c

QJ i-t
CL
O •
o0
1-
01 J*

QJ QJ
CL CL
O O
00 00
i. t
QJ QJ

OJ O"» OJ OJ
i— < ,— t *— t
S- S- t.

1_
QJ
C
re
CO CO CO
QJ QJ QJ
Q. CL Cv
O O O
00 00 00
t/i tn t/i
I- I- I-
QJ QJ QJ
•O TJ T3
C C C
re re re
OO OO OO
c:
o
o
                                                                                                      o o
                                                                                                      IT) UT)
                                                                                                      O O
                                                                                   S
                                  o «-< o in
                                  co ^H o o
                            00

                            o o
) CT> O IO 1^* CM r-.
> t-i ro to o •-* •-*
3 O OJ O i—(*—( O
                                        OOC5OOOOOO
                            888
                            O O O
                                                                                                                                                                      o
                                                                                                                                                                       V
                                                                                                                                                                      CO
                                                                                                                           .-i r«* cvj CO cxj    i-t
                                                                                                                           *t CNJ CVJ PO O    *^
                                                                                                                           O O O O O    •-<
                                                                                                                           O O O O O
   O
OO

00
                                        O

                                        O
CO cv r-. O O
r»- ^- ^ ^- O
O O O O •-«

o o' o* o o
                                                            O O O
                                                            000

                                                            O O O
                                                       o
                                                       ro

                                                       o
                                                                    o o o o
                                                                                                                                 O c\j
                                                                                                                                 n ^1
                                                                                                                                 C5 O

                                                                                                                                 O O
QJ
in "O

Ct- E v>
o •— c
^ re £
Sin o
VI
U (/> QJ
tft 4-» »*^
•«- Q.
E O re
0 £ U
O. O S-
^i iEc?
Z)
4->
i — t/l

fcl
l/l O
X 'f~
QJ t/1 >
•— 3 QJ
£ s*~ ^
Srel«
si §^
OO CL^E QJ
CCOOO<
un
OJ
trt t/l fO
3 -^ -f-
4-> J- -0
"o 3 "^ "re
tn u in c
re re QJ
QJ O)
t/l l/l C 4->
|l ||
fc t ' — O
CD CD CU 0
re
o
c
t- re
o u
••- M-
u re
CL
u •*-
t- l-
00 C OJ
i— o c
O i- O
UJ QJ i-
z £ <
re
-C
u
t/)
.c >>
u »
4-j re
3 ^:
Wl U
^ ^
t/» t/i
2£
M^
ft
I (J U
•—00



QJ
C
1_
cn
I

00


H^"5
I- 1- ^
3 CL4-*
in QJ
3 r— «
•r- re •»-

«.I g
CJ Q. Q.
m
QJ 3
ZJ '•-
o
in V-
4-> E
in
O wi
sl
in Q.
CJ ^J


re
§
i_
oo re
LU re
CJ -r-

re
re
c
CL
0
t/i
-*-> C

U •*-> t- X)
X u Zl
QJ tn in o m
• — 3 re re QJ
3 •—**-•—•«-»
CL re QJ
re CL zi 3 o
••- QJ I- U E
CL § 1 E •—
Q 00 CD O CL.



re
c
«/l
QJ
U
QJ
C
o
o
o
re

«
re
arcell
oo c:
h- 0
UJ QJ
Z Q-
re
o
c
u re
0 t/>
<«- o
re ,0
<_> re
i- C
re QJ
O t/l
L. re
QJ re
Q. O
si
o
3
J* QJ
o re
li
^ o • —
>-• O 00
w

•^ 3
C *O
o u
•O "*- Z»
4_i re
3 3 m
L. C 3
1 — in
O QJ */l
E > re

oo oo o
                                                                    •g

                                                                 00
                                                                 _) -t-
                                                                 X 4->
                                                                 O O
                                                                 
-------
0 0


OJ QJ
vi re to
|«|

.?
<
i
«e
Of
z:
0 OOO
01 10 0101 01
0>
Ol O) O) OJ
.££ .?.?.?
(J 1C 1C O O (J
Z! Lf) Z! £ ZT
4-> ID
o3 QJ CTt o3 o0 o2)
rH
z: z; u z; z: z:


UDUD
l£> Sen
t-H  to


CO
CT> VO
CL
(_> r-
00
•M
(/I QJ
^ QJ
to z:
o o
en en

QJ QJ
(/i t/>
QJ 0)
££
oo
0101

o> 

OJ d>
i/i to
M- -«-O
O U
£ z: en
^ J^ QJ
OJ QJ -^
z: z: a.
0 0
CP>  O 0)
^ z: LLJ :E
o o
CT> CT»
r-tt-4
QJ QJ
v> to
Od.— irvc
                                                                                                                               o«-ir-Hf-  Q.
             >>•!- (j U  QJ
            0 OL 1-1 •— —1
                                              en

                                           gg
                                           LJJ re

                                           < c
LOU
J —  OJ
                                                      i— 

                                                      3 3 '
                                                      S- T3
 X    ••- (J
 QJ 10 > in
*— 3 QJ re
   ar— t- l*_
   F- ^ •(- fC
                                                                                       •— to C  Q.-C
                                                                                       QJ t/i-i-  o CL
                                                                                       > re i-  i_  E
                                                                                       re *c >-i o ••-
                                                                                       1/1 tj cj z: a.
                                                                                                 tn i-  :D *•-
                                                                                                 3 •>- jr  o
                                                                                                 •— x:  ii E «n
                                                                                                 •— o  O ••" c
                                                                                                  o

                                                                                                 >> 
i §  e & «
CL CL O t- O
   SO  (X O 1-
   
-------
CQ
<
                O  O
                o o

                •« 00 *
                 .   .
                 s- t-
 cu  cu  cu  cu 4—1 cu cu
"O "U ^ "O    "O "CJ
 c  c  c  c o c c
 TO  TO  TO  TO O" TO- TO
CO CO CO CO 3 CO OO
     .   _         -
w-1  
f
0
o



CQ

o
ir>
CJ
	 i
t.
ID

O
in
O
_i
t.
^







co in
i-H OO
88
00
ID
r-.
o
o
OO
ro o







o
i—l CM O
§•-1 CO
0.-H
o o o



CO
8
d

i— « O
8S
d d




§Q
CM
dd
o
i
CO
r~t CM
88
do




8
O
d
                8
   o o
   o d
                                                                  o
                                                                  o
                                                                                                 «— 1 1— • cu i— <
                                                                                                               c
                                                                                                               O
                                                                                                           CTi   4-> C 4->  C
O  I-  C  C I-H C O"
    CU  O  O   TO  I-  V. i"H t-
o  3t  eu  cu     cu TO
CU    T3 T3  M -O C
**-  C  C  C 4-> C 1-

LJJ  *" z x i<; z £
                           .  _   .      -  -
                          QJ  CU   O     *D CVJ CO VO CM
                                                                                                           OOO     O ^H O O O

                                                                                                           ooo     odooo
                         n ^-» o o •-< O
                         o o o o o o
                         o o o o o o

                         o o o o o o
                                                                                                                                                 O in       —i
                                                                                                                                                 o o       o
                                                                                                                                                 o o       o
•D
ro .-«

00

0*0'





CO
3
C_5
^
J—
CO
s
in
3
4->
11
ss
«J X •«-
C Q)  >
il.|I
iomo.cn
•t- -t- QJ 3
C C CJ'-
.C £ O »—
O- CX E CU
Q O OO  cu m c
H- »— cx+-> in
cu cu
irt in in c 4->
3 3 3 O O


CD CD CD O- O
C *O
cu o
3 C
tr -i- o
••- -o <*-
3 ^3 ^
J- ITJ

in i —
tcu >i
u u
TO fe
co tyt c c
f— _0 0 0
CU JT CU CU
CO Q.4-* 4-» 4
•Z. O Q- Q. C


0 -Q O>

»— in
, >
3 CU in

C « °
£ w-o
0 G **

S-
u
o
t.
E
in


I CX
CO CU






UJ

<
ft
ce




c
I*
re

c:
^
CX
o
in
*J
u S
4-> «
3 -C
in O
CU

III
c c m
>, >> en
^ x:
i- S- 0
o o E
x u u •—
co c c: «o
•— 0 OCO

in^
3 ^
J^ £
13 CX
*n cu

in _c
in a.
fO •!-
CJ 0-

*TJ
3
(J

•*->
l-
.|2


^
o
0.
3
4-1
S3
3 >r- in
u ^: c
«3 U CU
O L)
in u «n
J 1 1
O m »*-

cu E TO
4-> O U
m ex i.
TO CU CU
CD _J CX.
3 3 m
4-> TO C
3 3 *n m E TO
J_ -0 3 "- -^ -r-
L 3 *J t- i — -O
in o •«- in -o .^
rO X ••- U 3 3 in
ccuin>in(j(U(/)'r-
Z E CX TO JD CU
<5 JC in ^n m c in
LUTOTOO-l/)333O CU
O ••-•!- CU31-1- l.E*->
-------
                          OOO    OOOOO
               XI O
               LD O"

               Si
J QJ  <1J    QJ Q) (L>  GJ


i vi  v* «— v» v* v»  \n
                                                          QJ  QJ


                                                       i-~ t/i  vi


ID tf)
CT» CTt
QJ QJ
O. Q.
O O


C\J
r-.
c_)
3
S£
77
o o
o0 o0
TJ -D
C C
CO LO


f1** to
i— 1 1— i
VI tO
QJ QJ
C C
CO CO


C\J
to
QJ
c
CO
rv.
CTi
03
QJ
J*
QJ

_E
Q_
^
+J
VI
1
J*
(J
£
rx i-
cn 4->
.-H VI

00 QJ
•s«
U O
z: z:
t.
CO
1
•0
.^
QJ
U
^
i-
V) CTi
•i- i— I
+->
a0 QJ
JK: -^
QJ QJ
O U
 t\J *3- .—I
                                                                                   c\j on o
                                                                                   .-H CVJ ^H
                                                                                   o o t-<

                                                                                   ooo
                                                                           CVJ O CM
                                                                           O ro.-f
                                                                           O^H O
CD
                        CO ID


                        CT> CO
                                                                          CO
                                                                                   §r-. r--
                                                                                   .—t m
                                                                          o       o o
                                                                     vo ro
                                                                     t~( 
1
?ji

: o
. c

«
Q. t—

x:

vt
sfc
VI C
X ^
c'w

x:
z 

v* v. nj V.

'JjJ ^^ .*Z
•^ O «5 (J  (J
*— 3 (O *D
|^«^^

CJ -i- Q) 1_ i.

O£ O CO t3 CD
1C
n U
C
o

U
U

K 0

Z Q_

X

vi
QJ
VI C
(j ••-

= 11

.— OLO



v>
S- Q.
^'31-
!' * «

Cl VI •!-

 V>

•— l_
£ VI X
1. f— O
O. QJ i.
e D. (J
VI *w
QJ v» E

4| E
E ** O.
•i" U QJ
Q. l-l ^- _l
             -  -
             >- C



             1S


               "
                        fe*

                        So
                                    l«


                                 S'S^
                                 Q_ (J '-
                                 Z QJ J-
                                 i-« vt «J
                                                                                               t— ••-    U I-
                                                                              150
-------
                             QJ OJ
                             CL CX
                             O O O
                                  *

                             I/) (/>  C  Ol
                           o; i- *j
                              o  c
                           T3 -Q  Q)
                           C C  E
                              c i—i c
                                   §OJ QJ
                                O E Q-
 O
 U
                                 O VO
                                 §§
CO
O 00

o o

o o
CO      t-<
°      CD

O      0
                                               o
                                               o
                                                             o ^r o o
                                                             CO ^H ^H C"> O
                                                             CM o ^H o oo
                                                                                    O t-H      ^H
                                                                                                            o       ^-«

VI



•^ 2
CL E
o •—
•— re
p vi
O VI
i i.

it
O. 
cr
to, ••-
3 V)
*-> c
m oj
i — -^
3 Vt v>  I- -Q
QJ tO 4-» re
vi f~ »/> j«r
X U 3
QJ V> V> U VI
•— ^ *O *D QJ
=»•—*»- — *J
CL re oj
JC VI VI C
*TJ CL 3 3 O
••- QJ i- t- E
C U "3 fl QJ
-c o £ E re
CL E C E »—
re ••- re re re
Q t^ CD 15 ta-
re
u

C
1_
0
<^_


re
o
v>
>>
U
J-
re
10 C
1- 0
t_> i-
LU QJ
LO +J
SZ Q.


-C
(_) V)

3 -4-J
vi re
•r- •*-»
^ O V>
C 3
VI 3 >~
3 D. re
^ -c:
u vi CL
C 3 QJ

•£3^
Z 8 -2 01

«— ' O •— « Z








1_
Of
c:
T3
1_
re
C7>
§
X —
to re
i-, to




^
i_

^
u
«£
SS
3 E
1.
4-> U)
ii
»= CL
re QJ
to _i
*/> •«-
3 V)
4-> C
ro OJ

3 3 vi v re

S- 3 4-> i- TD
Qj re re 4_> re 1/1
t/i o •*— t/> ^ •!—
x ---03 re
OJ vi > vi u vi c
i— 13 aj re re oj
to 3- — L»*-»— -t-> v>
Z Q. 1C ^ QJ QJ
^ .C V) V> C •*->

O-i~O)=3S-l_EQJ
•fctjf— re«ojc
1— ^o— E.E.reo
tocLEojEfc'— u
ZD re *«— vi re re re t-
O£QtO«tCJtDQ-O
re
u

c
i_
O
4_-


re
o
VI
>,
o
i.
re
to c
»- 0
(_> i-
UJ (U
tO 4->
Z Q-



VI
3
J_

_c
u
o
1-
u
re
E
V)

i CL
on QJ










re x
C OJ
to re 3
5EQ
uj re m
«t c c
I— JC -C
to CL CL

or O o (/
/i
3



VI



VI
3
u
re
3
s_
re
j £


OLD
       •g
                                                      •g
                                                    §5
                                                    Sf
                                                                         151
-------


en
ID u"i
25;

QJ -i-
c re
to CL
CO CO


CM
VO
C^
*— t

4->
a) in
— »
c: i— i ^o
0) — ,-i
o re cj
••- CLO-
CL. CO 3


CM CM
IT) VD
c* 
• t— I •
re +j re

4->
QJ in
r*.
CZ i-H
OJ 'I-
o re
•i- CL
Q- CO



QJ -M
c Q*

jj C C
Q) QJ
l~ L> U
o ••- **-
^ CL. Q.


VO
to
Cft
OJ
CL
o


0)
£


ID ^Ct
QJ O)
ex Q.
O O CM
p** Cft
1— 1
•r- U 0) QJ


O1 CM CM
£. cn CT* cn

T, £££
cxrererecxrerere
COCOCOCOl/lCOCOCO

i— i

r*.
O




CSJ CM
r-H «— »
ss
VD
in CM
O CVJ
30 *D
i— I
C
QJ-,-
CL»4-
O 3

^-a
QJ CIJ
JS

CM
CO
OJ
ex
o
o

CM CM
—t tUr-t
0 C 0
o* re o*
3t to 3
CO
O O
c
•*-
3
re
til
oa
c
(U
c
QJ

CO
CO
Cn ID
•— i
CL *
O •—
r—
CM Q) i— 1
1- CT* C OJ
QJ ^H -r- -r-
C O 3   d)  a>
                                                                                                                                                  •o -a -a -o T3
                                                                                     00    00
                                                                                       II     t  I
                                                                                     o o    o o
                                                                                     in in    in in
                                                                                     CO C_>    CJ CO
                               o o in o **• «-H

                               o O o o O o
                                                                 8
1O CM
•-H O
o o i

O o r
                   CO


                   i-H    O
                         «-H       O
                                                                                                                                                  qq
                                                                                                                                                  o o
                                Sre  i.
                          VI    *— -i—
               _
              ••- 3 
                 -
                     :  3 V)    *— -r-
re ui
4J 3
re i.

u u
•i- o
-M 5-
QJ U
s- re
E
re
•i- irt

QJ Q.
0 	 1

t/l
^
4J
re


2
O
1-
o

t/1
3
QJ
C
re


1
&

i/i
£
re
Wl
13

a
E
u
•E


re
c
en
to re
§e
LU re
- V) _^
X -^- O 3 VI
QJ I/I > V) O V) '*-
31 — t- H~ i — 4-> C
ex re j3 QJ
.C l/> (/I CZ l/>
re ex  3 3 o QJ
•«- QJ 3 t- I- E •*->
c o r— re re QJ o
ex e 'cu 1 1 •— c
re -r- vi re re re s-
O CO 
in
re
QJ
S.
CO QJ
I- I
CO QJ
UJ JZ
co ex
Z UJ
insulen
CL
i-
i/l
CL
E
O
01
0
1.
0
re
u
c
1- fl
O 4-
"-
re
o
V)
>i
u
re
0
0)
CL.
3
j re
"D
JD
re
OJ
u

c
4J
CL.
vi re
re o
o ^*~
(J o
>, re
•— ex
re re
s_ s_
3 3
QJ QJ
C C
O O
i- u
U 1-
«r <
(O •!-
VI C
o o
r~~ 4J
js "aJ
re jn
^ QJ
re jr
c: >>
QJ V)
v> CL
S£
re -a
(O Z
re is 3
«~ 4J 1-
QJ UJ •<-
•*- E ** jr
J_ O U O
QJ i~ C O
C CL 3 1-
•0 CL O
i- i/i re
••- QJ vi E
re i — 3
01 re t_ wi
£ 3 ••-
§0.-— E
 ex
co re •— u QJ
i-i CO CX «-• _l
3 (rt Ml
fc5I
tl) « •*-» ui
Wl •!- l/» -f-
x u 3 re
QJ V) wi (J C
' — 3 re re
CO 3 « — >*- < — */»
z CL re QJ
<, £ t/t (/) 4->
uj re o. 3 3 u
C_J ••- CU 1_ I- QJ
< c o re « c
t— -c o £ £ o
i^ ex E E E u
^ re •»- re re i_
C£ O CO ID CO O
                                                                               152
-------
               CO       r-. r
                                                                              3 (_j      r- r- r- ,—
                                                                                                          O> OJ
                                                                                                          Q. O.
                                                                                                          O O CT>
                                                                       I/) I/I (/)
                        ••- o
                         <_> Ul
                         (O -^
 c
 o
 o
                                                                                        O •-» O O

                                                                                        O O O O
                                                                                                                CVJ      •—<
CQ
                                                                                                          LT) CM CO
                                                                                                          CM •* ^H
                                                                                                          O O O

                                                                                                         'o o o
CQ
<:
                (O       f- fc. >O
            CO  C       (/> 3 t-
            I—  O       V) •— O
            (_>  S-       it) rtJ ^
            UJ  HI    OC 1- 4-» V)
            l/l •*->    CO «J O (d
            zo.    ^-i o —'a:
a- g  c
                                                                          f—  t/»  t/>
                                                                       X  U  3
                                                                       di  i/t  o  ->-
                                                                                     a. a.
                                                                                 153
-------
                •—        (1)
                                                                                                OOOOOO
                                                                                                r- r- i-- r-~ r- t-^
                                                                                                                                 CL       ,-t

                                                                                                "O T3 "O TD T3 "O
                                                                                                 C  C  C  C C C


                                                                                                OO OO l/l OO OO LO
                                                                                                   OJ -Q
                                                                                                   CL  It)
                                                                                                   O r-
                                                                                                  o <
                                                                                                CQ CQ CO CD CD _J *—- CO
                                                                                                CD CD CD O tS    LU CD
                                                                                                Cl_ Q_ Q- Q_ Q_  V- CD Q-
 c
 O
 (J
CQ
CQ

t—I t-H IT)       -—I

O O PO       ^H
                                                                                                ^-   -000

                                                                                                CSJ   *•  »t—t OO <
                                                                                                  -o c
                                                                                                  50(
(_) O
UJ UJ


o o-
us kisutch
_c
o
c
>)
JC
0
O)
c
•o




h0
i-l
auratus
promelas
punctatus
V)
(rt 3 1/1
3 r- 3

in O,-—
to a; id
b^s
(_J CL. ^^
eticulata
crochirus
U R>
E
(0

•^ i
Q 	 1
X

ECU

re)
<'c
00 Q. 1
:D w
c£ a u
i
3
I/I
4*>
I/)
u




I- OO
«o t—
i a
-S3 oo
0 O Z
s californica

u

fO
c
0
l_
4-*
O-
us tschawytsch
dneri

u «-
c «

JZ
i- 0
0 E
a: o •—
•— i O OO
••-So
O t- i_
(U CL (J

a>


CLjC E
o a. o
s- a) CL
•*-* E O
O ••- 
(_) f— ^3 3
«t c-.-~
t/i'S. *" "
es kadiakensis
nais


                                                                                                                                         )  c c  ia
                                                                                                                                         « O O 00 Q_ _| QC
                 §-*-> u
                 (J -^
              Z OJ t-
                                                                                    154
-------
                                                                                                                   (/> in in
QJ QJ QJ QJ QJ QJ QJ
C C C C C C C
CO tO CO CO CO CO CO
en
en
•-I •-! OJ

c c 
                                                                                                      fo    r-v .~i
                                                                                                      .-H    O ^-i
                                                                                                      o    o o

                                                                                                      O    CDO
                      3  in  ui  B3
                   (D -O  3 T !T!
                            1- -O
                           4->  «J  i/)
                      o -»-  m -^ •»-
                     ••-03     *0
                      > in  u  w  c
                      QJ 10  nj  n;
2    g
                                                -M       »—
: VI JD       Qj QJ
  d    i/i in c 4->
5 O */> 3 3 O U
- "O 3 1- C, E OJ
: ••- r- *o « QJ c

x e.'oi E E »— u   :c
J >i VI »O —« -
- j: o
i Q. E
3 ID -r-
-r-  (J

Ts

I«

u i—

CJ <
                                                           QJ
                                                          J*
                                                        l/i  »O
                                                        3 •*-
                                                        4-> -o
                                                         -t-
                                                                                                                     E -
                                                                                       SO -i
                                                                                       •r- i
                                                                                   155
-------
                      ooo    oo


                      i— 1 i-H i— 1    i— I«-H    LD
             0) in    QJ  (U  a; r--  4->
                                     r—  l/l VI VI
                                                   IT) «— 1
                                                   01
                                                   .-1  1-
                                             c

                                         -M O i
VD IO
tO VO
en en
 O
J CO
a o0

/> vi

3 "O
CO
10
en
OJ
0.
o
o
00

VI
QJ
T3
zee
Q re re
O CO OO


XI
in
ID
en
t-H

QJ
CL
3
ID ID
cn cn
r-t r-4
re re
-*j •*->
OJ QJ
cn cn
c: c
j_ i_
QJ QJ
^t -^
U (J
cZ cT
2
re
4->
QJ
JD
cnin
C VD
•r- CTi
t. I— 1
QJ
J* OJ
(J CL
•i— o
CL. C_)
vO
cn
re
+j
OJ
Cfl
c
t.
QJ
_^
O
Q.
in in vc>
*D vo vr>
cn cn o>
•—(«—<*-*
0)
CL
O
<_>
00

VI
t-
QJ
•o
QJ O>
&CL
0
C_J CJ
00 00

VI VI
S- i-
QJ QJ
•0 T3
i£>
ID
cn
QJ
CL
ss
-2

VI l/l
1_ l_
QJ QJ
-OT3
C C C C C
re re re re re
CO OO OO OO CO
^>
CO CO VO
«D  '
cn cn ^H
ai 0    00
) ro ro OJ oj
!> O i-H O t-l
> 00 OO
                                    oo    oooooo
                                                                              ooo
                                                                              ooo
O ro ro
ro OJ «—<

888
000
                                                                                                                                                    s
                                                                                                              t-H o o
                                                                       o o
                                                                       oo
                                                                                 o
                                                                                 r-
                                                                                 O
                                                                                             3 O O
                                                                                             3 O ^C
                                                                                             3O O
                                                                                                                                                i—I ID ID CJ
                                                                                                                                                ,-< ro •-! od
                                                                                                                                                o o o o
                                                                                                                                                CM O    O
                                                                                                                                                CNJ co    r*-
                                                                                                                                                01-<    O
                                                                                              Ot CO *£)
                                                                                              O «-H O
                                                                                              ooo
                                                                                                                        cvj in CD
                                                                                                                                                 OO    O
                                                                                                                                                 ITJ CVJ    •-•
                                                                                                                                                 Of    *—'

                                                                                                                                                 00    O
re
c

0
re M-
4->
4-> l/l
3 3
i5
II
re re
CO CO

u
VI
3
•M
re,
4-1
(_J
C
3
CL
VI
3
t-
3
«J
4->
u
ulata
culeatus
 (U
2S
Ol »—
4-> «4-
&«
i- o
O J-
£^


OO
UJ
?
CO
s


X
QJ
1
re
c
si
cx
re
o
VI
3
4->
re
Is
i- •*->
QJ re
l/l •!-
u
VI l/l
3 re
• — i»-
<0
JZ V>
CL 3
,tris
adiakensi!
vi _a£
3
(J VI
re QJ
< — 4->
0)
VI C
? S
Q) 1. S- fc
u n> re QJ
OO C3 CO Ou
                                                                                                                                                               re i-
                                                                                                                                                               ••- O
                                                                        i—    UJ  QJ  O> I
                                                                                                                   =  a; •*"
                                                                                                               •^ 4->  E ^
                                                                                                               i_  re  o o
                                                                                                               a)  i-  i- o
                                                             £= rO  O)  O
                                                         re •— t  E  CL
                                                         i>o re re -i-  QJ
                                                         i—. LO <_J Q_ 	I
                                                                                                                                         1_ C7H—  3  ro
                                                                                                                                       I  10 03 3 r— i—
                                                                                                                                       :  u e cx re
                                                                                                                                     r—       -
                                                                                                                                     C/l CX Q. D. E  E
                                                                                                                                     3 ro *a  »O ••-  
-------
                    •»-> 0) IO
                    0)r= Cft
                                       CT»

                                       1-1 CL CL
                                        • O OOOOOO^OO
                                       r— <_)(_)!— r-. i*-. r- *o r-» r-^
                                            TJT3'OT:'O"
                                            ccccc
                                                                                                    o o o o ty\ o o
                                                                                                    rx. r-. p— r*» \o r»* r*.
                                                                                                     •O -O -O "O TD T3 T3
                                                                                                      c  c  c  C  c  C  c

                                                                                                     (/I IO LO IO CO CO IO
C
O
O
              o 
                        «er o o
CQ

LU
_l
CQ
                          CM Cft
                          00

                          00
m    o o
o> 10 ^j- in m

OOOOOCMt—(LOt—IO
                           m
                                                                                                    •—« o in oo in r
              O CM

              o o
                                                        CM CO

                                                        ro co
                    •— -4-* t.
                    m to -r-

              t-    O U  ••-
>

(/)
1

0


^
t.

3
(J
to
V>
1

5
••-
I
5


»0
(/> C
O)
at at
c 1
E -u to
i — u LU
 >
«!
O, t/i
|3
1- O)
CL Ul


l^ 1/1


O 3
t/1 O
{/I (O
3 3
fete

55
                                                                                  157
-------
                                              to
                                                     10
                                                     3
                                               c     o
                                              •r-     01
                                               S_    -i-
                                               Q-    e
                                                     to
                                               to          0)
                                               o     E
                                                                            s-
                                                                            o
                                                                            Q.

                                                                            s_

                                                                            to
                                                                            T3
                                                                            O)
                                                                            to
 o

-(->
 S-



 o

 CJ
•r—

 O)
                                                                             CD

                                                                             O
                                                     to
                                                     3
                                                     C
                                               H3     >
                                              CO     r—
                                                     co
                                                                            CQ
                                                                             OJ
00
CO
-------
                       o>
                       (O
                      -o
                                               o
                                               o
                                              0)
                                              S-
                                              CD
co
oo
et
          O
          o:
          o
                           159
-------
TABLE  B.2.   SCIENTIFIC AND COMMON  NAMES  OF  FISH USED  IN  THIS REPORT.
                Scientific Name                            Common Name
      Oncorhynchus kisutch
      Oncorhynchus tschawytscha
      Salmo clarki
      Salmo gairdneri
      Salmo trutta
      Salvelinus fontinalis
      Salvelinus namaycush
      Esox lucius
      Carassius auratus
      Cyprinus carpio
      Notemigonus crysoleucas
      Notropis atherinoides
      Notropis lutrensis
      Notropis umbratilus
      Pimephales notatus
      Pimephales promelas
      Rasbora  heteromorpha
      Catostomus commersoni
      Ictalurus melas
      Ictalurus natal is
      Ictalurus nebulosus
      Ictalurus punctatus
      Gambusia affinis
      Poecilia latipinna
      Poecilia reticulata
      Gasterosteus aculeatus
      Lepomis  cyanellus
      Leponns  gibbosus
      Lepomis  macrochirus
      Lepomis microlophus
      Micropterus dolomieui
      Micropterus salmoides
      Perca flavescens
      Stizostedion vitreutn
      Mugil cephalus
      Morone saxatilis
coho salmon
chinook
cutthroat trout
rainbow trout
brown trout
brook trout
lake trout
northern pike
goldfish
carp
golden shiner
emerald shiner
red shiner
redfin shiner
bluntnose minnow
fathead minnow
harlequin
white sucker
black bullhead
yellow bullhead
brown bullhead
channel catfish
mosquito fish
sailfin molly
guppy
threespine stickleback
green sunfish
pumpkinseed fish
bluefill sunfish
redear sunfish
smallmouth bass
largemouth bass
yellow perch
walleye
striped mullet
striped bass
                                            160
-------
ORGANOCHLORINES
CHLORINATED HYDROCARBONS
Aldrin
Chlordane
DDT
Heptachlor
Isodrin
Lindane
Mirex
TDE (ODD)
Toxaphene


.011-. 2
Very low
.01-. 037
.056

10
.001
ins
1.5
OXYGENATED-CHLORINATED HYDROCARBON
Chlordecone (Kepone®)
Dichlone
Dicofol (Kel thane®)
Dieldrin
Endosulfan
Endrin
Methoxychlor
Tetradif Ion
ORGANOPHOSPHATES
ALIPHATIC DERIVATIVES
Acephate
Demeton
Dichlorovos
Dicrotophos
Dimethoate
Dioxathion
Disulfoton
Ethion
Ethoprop
Malathion
Methamidophos
Mevinophos
Monocrotophos
Naled
Oxydemeton -Met hy 1
Phorate
Phosphamidon
TEPP
Trichlorfon
HETEROCYCLIC DERIVATIVES
Azinphos-Ethyl
Azinphos-Methyl
Chlorofenvinphos
Chlorpyrifos
Coumaphos
Diazinon
Methidathion
Phosalone
Phosmet
Thionazin
1.5 to 2.0
.1
ins
.186
<1
.23
.62
200 at 50°C


very
6.6
1.0%
misc
2-3%
ins
25
1
ins
145
9%
misc
misc
slightly
300
50
misc
misc
13%

ins
33
145
20
1.5
40
240
ins
25
1140
TABLE  B.3.   PESTICIDE  CLASSES

 Solubility (mq/1)       Kd
                    5 x 103

                    5 x 103
                        104
                        104
                        103
1 x
1 x
5 x

1 x
1 x
5 x
                        104
                        104
                        104
                    25
                    5 x 102
                    50
                    50

                    25    9
                    1 x 102
                    1 x 102
                    5 x 102
                    50
                    1 x 102

                    10
                    50
                    25
                    50
                    1 x 102
                    5 x 102
                    50
                    50
                    10
                    1 x
                    1 x
                    5 x
                    50
    102
    102
    102
                    50
                    50
                    5 x 102
                    5 x 102
                Soil
              Persistence
              very  stable
T/2 :.-7 yr
T/2 4-8 yr
T/2 200 d
              T/2  122 d

              T/2  290 d


              1  wk
              T/2  26  d
              2  wk
T/2 140 d



T/2 20 d

T/2 29 d

12 wks
                  Water
                 Persistence
7

5
1
1
7
1
5
5
5
x

X
X
X
X
X
X
X
X
10*

104
105
104
104
103
104
104
104
T/2 1-4 yr

T/2 2-4 yr
T/2 3-10 yr
T/2 7-10 yr
T/2 4-8 yr
1 yr
very stable

4-16 yr
40%
20%
85%
100%
25%




9-10
- 4 wks
- 8 wks
- 8 wks
- 4 wks
- 2 wks




yr
high at 1  ppb
or less

high
100% - 8 wks
30% - 2 wks
5% - 4 wks
100% - 8 wks
20 - 38 wks
                62  d  at  20°C

                85% - 4  wks
                10-25%  -  2 wks
                0-10% - 4 wks
                             100% - 4 wks
                             T/2 7-8 hr
                             detected at 256 d
T/2 30 d at  pH  9
                                                 50% - 1 wk
                 161
-------
PHENYL DERIVATIVES
    Carbophenothion
    Chlorothion®
    Crotoxyphos
    Crufomate
    Dicapthon
    EPN
    Famphor
    Fenthion

    Fonofos
    Parathion

    Parathion-Methyl

    Ronnel.
    Stirofos
    Temephos (Abate®)

PHENOALKANOATES
TABLE B

Solubility (mq/ll
2
40
110
ins
35
100 - 11%
slightly
55

13
20-24

50-60

.3. (contd)
Soil
1 Kd Persistence
1 x 103
2 x 102 T/2 - 36 d
1 x 102
1 x 102
5 x 102
1 x 103
2 x 102
5 x 102

2 x 102
5 x 102 7-20 d

3 x 102 2-60 d


Water
Persistence







50% - 2 wks
10% - 4 wks

50% - 2 wks
303! - 4 wks
11% - 2 wks
-0% - 4 wks
40
11
ins
ALIPHATIC ACIDS AND ESTERS
    Dalapon                  50%
    Glyphosate               1120 mg/1
    Trichloracetic acid (TCA)83%

AROMATIC ACIDS AND ESTERS
    Bifenox                  .35
    Chloramben               700
    Chlorthal-Dimethyl (DCPA).5
    Dicatnba                  4500-7918
    Endothall                21%
    Fenac
    Naptalam
    Picloram
    Propargite

PHENOXY COMPOUNDS
    2,4-D
    Dinoseb
    Erbon
    MCPA
    Si 1 vex
    2,4,5-7

PHENYLAMIDES

AMINES AND ANILINES
    Alachlor
    Bensulide
    CDAA
    Diphenamid
    Pronamide
    Propachlor
    Propanil

NITROANILINES
    Benfluralin
    Butralin
    Dinitramine
    Fluchloralin
200
200
430
ins
620 at 25°C
52
ins
27%
140
140
240
25-50
2%
260
15
700
500
 .5

 1
-70
2 x 102
2 x 102
                 0.2
                 0.2
                 0.2
0.5-1
0.5-1
0.5-1
0.5-1
0.2

0.5-1
0.5-1
0.5-1
0.5-1
1.0
10    ,
1 x 102
1.0
2.0
2.0
50
5
    102
1 x
50
50
10
5 x 102
5 x 102
50
1 x 102
              15-30 d
              150 d
              20-70 d
                               40-60 d
                               40-60 d
                               400 d
                               2 months
              350-700 d
              20-60 d
              550 d
              10-30 d
              15-30 d

              30-180 d

              5 months
40-70 d
10 months
20-40 d
90-180 d
60-270 d
30-50 d
1-3 d
              120-150 d
              90-120 d
              90-120 d
                              2-3 d
                                                50%
                                                10%
                                    1 wk
                                    2 wks
                                                1%  in  30  d
                                             162
-------
                                TABLE  B.3.    (contd)
    Nitralin
    Phenoxalin
    Prof1ura1 in
    Trifluralin

UREAS
    Chlorbrorouron
    Chloroxuron
    Diuron
    Fenuron

    Fluometuron
    Linuron

    Monuron
CARBAMATES

    Aminocarb

    Benomyl
    Bufencarb
    Carbaryl

    Carbofuran
    Chlorpropham
    Karbutilate
    Methiocarb
    Methomyl
    Mexacarbate

    Propham
    Propoxur
THIOCARBAMATES
    Butyl ate
    EPIC
    Molinate
    Thiram
    Vernolate

TRIAZINES AND TRIAZOLES

    Ametryne
    Amitrole
    Atrazine
    Cyanazine
    Metribuzin
    Prometon
    Propazine
    Simazine

MISCELLANEOUS
Solubility (mg/1)       Kd

   .6               50
   .5               1 x
   .1               5 x
   1-24             5 x
    103
    102
    102
2 x
50
3.7
42
3850

90               20
75               1 x

230              50
                     102
                 5 x 102
                 1 x 102
                 10
    102
slightly         1 x 102

ins
100
40-99

700
88-108
325
ins
5.8%
100

250
2000
15-300           5 x 102
370              1 x 102
800              50
30               5 x 102
90               1 x 102
185              8
28% at 25'C      1
33 at 27°C       5
160-171          3
1200             1
677-750          8
8.6              1
5                6
                                 Soil
                               Persistence

                               moderate

                               320-640 d
                               120-180 m
              300-400 d
              200-500 d
              30-270 d
                               120 d

                               1250-350 d
                                                                              Water
                                                                             Persistence
              40-80 d
              30 d
              80 d

              50 d
              30-90 d
              15-30 d
              300-500 d

              150-200 d
              >400 d
              200-400 d
              200-400 d
                                                  3 months
                                                  60* -  2  wks
                                                  20% -  4  wks

                                                  20* -  4  wks
                                                  none - 8 wks
                                                  40* -  2  wks
                                                  30* -  4  wks
                              60% - 2 wks
                              10% - 4 wks
10
5 x 102
5 x 102

5 x 102
1 x 102
2
1 x 102
5
1 x 102

50
1 x 102


16 wks
T/2 8 days


120-260 d





20-60 d




5% -
0% -





15%
0% -

50%
30%


2 wks
4 wks





- 2 wks
4 wks

- 2 wks
- 4 wks
BOTANICALS
    Allethrin                ins
    Pyrethrum/               ins
      Pyrethrin (synthetic)
    Rotenone                 slight
1 x ID*
1 x 104

1 x 103
                                               moderate

                                               decomposes  readily
                                           163
-------
ARSENICALS
    CMA
    DSMA
    MSMA

DIAZINES-URACILS
    Bentagon
    Bromacil
    Pyrazon
    Terbacil

DIPYRIOINUMS
    Diquat

    Paraquat
CYANATES
    Chlorothalonil
    Dichlobenil

METALOIDS
    Copper Napthalate
    Mancozeb

OTHER
    Acrolein
    Captan
    Carboxin
    Difolatan
    Dodine
    Methazole
    Methyl Bromide
    Norflurazon
TABLE B.
Solubility (mg/1)
2%
2.8%
3. (contd)
Soil
Kd Persistence
0.2
0.2
0.2
Water
Persistence

500
815
300-400
710
7%

completely

0.2
5
30
50
5 x

5 x





103

103


T/2 5-6 m
30-60 d
700 d
>500 d rapid
inactivation
>500 d
>2 yr in mud




7-27 d

7-14 d

.6
1.8
ins
moderate
>.5
170
;ns
6300
1.5
1.75*
28
50
20
                 5 x 102
                 5
0.2
30
10
50
0.5
5
10
50
60-180 d
2-3 m
                                             164
-------
REFERENCES

American Chemical Society.  1969.  Cleaning Our Environment:   The Chemical
Basis for Action.  Washington, DC.

American Fisheries Society.   1970.   A List of Common and Scientific Names
of Fishes from the United States and Canada.  R. M. Bailey (Chairman),
Third Edition, Special Publications, No. 6, Washington, DC.

Alabaster, J. S.  1969.  "Survival  of Fish in 164 Herbicides, Insecti-
cides, Fungicides, Wetting Agents and Miscellaneous Substances."  Interna-
tional Pest Control, March/April, pp 29-35.

Allison, D. T., and R. 0. Hermanutz.  1977.  Toxicity of Diazinon to Brook
Trout and Fathead Minnows.  EPA-600/3-77-060, U.S. Environmental Protec-
tion Agenty, Duluth, MN.

Battelle's Columbus Laboratories.  1971.  Water Quality Criteria Data Book
- Volume 3.  18050 GWV05/71.  U.S.  Environmental Protection  Agency,
Washington DC.

Bond, C. E., R. H. Lewis and J. L.  Fryer.  1960.  Toxicity of Various
Herbicidal Materials to Fish.  R. A. Taft.  San. Eng. Cen. Tech. Rept.
W60-3:96-l-l.

Bridges, W. R. and 0. B. Cope.  1965.  "The Relative Toxicities of Similar
Formulations of Pyrethrum and Rotenone to Fish and Immature Stoneflies."
Pyrethrum Post, 8:3-5.

Brungs, W. A., R. W. Carlson, W. B. Horning II, 0. H. McCormick, R. L.
Spehar and J. D. Yount.  1978.  "Effects of Pollution on Freshwater Fish."
Journal of Water Control Federation, 50(6) -.1582-1637.

Burdick, G. E., H. J. Dean,  and E.  0. Harris.  1964.  "Toxicity of Aqualin
to Fingerling Brown Trout and Bluegells."  N.Y. Fish and Game J. ,
Call, D., L. Brooke, and R. Kent.  1979.  Estimates of No-effect Concen-
tration of Selected Pesticides in Freshwater Organisms.  Third Quarterly
Progress Report to EPA, 1 January through 31 March 1979.  U.S. EPA Grant
No. R-806196-01.  University of Wisconsin, Superior, WI.

Cardwell, R. D., D. G. Foreman, T. R. Payne, D. J. Wilbur.  1977.  Acute
and Chronic Toxicity of Chlordane to Fish and Invertebrates.  EPA
600/3-77-019, U.S. Environmental Protection Agency, Duluth, MN.

Cairns Jr., J. and A. Scheier.  1964.  "The Effect Upon the Pumpkinseed
Sunfish Lepomis gibbosus (Linn.) of Chronic Exposure to Lethal and Sub-
lethal Concentrations of Dieldrin."  Notulae Natur, (370)1-10.

Carlson, a. R.  1972.  "Effects of Long-Term Exposure to Carbaryl (Sevin)
on Survival, Growth, and Reproduction of the Fathead Minnow, (Pimephales
promelas)."  J. Fisheries Research Board of Canada, 29 (5): 583-587.
                                      165
-------
Carlson, C. A.  1966.  "Effects of Three Organophosphorus Insecticides on
Immature Hexagenia and Hydropsyche of the Upper Mississippi River."
Trans. American Fisheries Society. 95(1):1-5.

Cope, 0. B.  1965.  "Contamination of the Fresh-Water Ecosystem by Pesti-
cides."  Journal of Applied Ecology, 3(Supplement):33-44.

Cope, 0. B.  1965.  "Sport Fishery Investigations."  In The Effects of
Pesticides on Fish and Wildlife."  U.S. Fish and Wildlife Service,
Circ. 226:51-63, U.S. Department of the Interior, Washington, DC.

Cope, 0. B.  1963.  "Sport Fishery Investigation."  Pesticide-Wildlife
Studies, A Review of Fish and Wildlife Service InyestigatJgns^)u~rJng 1961
and 1962.U.S. Fish and Wildlife Service, Circ. 167:26-42, U.S. Depart-
ment of the Interior, Washington, DC.

Crosby, D. G. and R. K. Tucker.  1966.  "Toxicity of Aquatic Herbicides to
Daphnia magna."  Science. 154:289-291.

Donigian Jr., A. S. and H. H. Davis Jr.  1977.  Agricultural Runoff Manage-
ment (ARM) Model User's Manual:  Versions I and ~Tl.Prepared for the U.S.
Environmental Protection Agency, Athens, GA.

Eaton, J. G.  1970.  "Chronic Malathion Toxicity to the Bluegill (Lepomis
macrochirus rafinesque."  Water Research, 4:673-684.

Environmental Protection Agency.  1972.  Effects of Pesticides in Water -
A Report to the States.  PBB222-320, Prepared for Office of Water
Resources Research by NTIS, U.S. Department of Commerce.

Environmental Protection Agency.  1976.  Quality Criteria for Water.  U.S.
Environmental Protection Agency, Washington, DC.

Ferguson, D. E., D. T. Gardner and A. L. Lindley.  "Toxicity of Dursban to
Three Species of Fish."  Mosquito News, 26(l):80-82.

Gaufin, A. R., L. D. Jensen, A. V. Nebeker, Thomas Nelson, and R. W. Teel.
1975.  "The Toxicity of Ten Organic Insecticides to Various Aquatic Inver-
tebrates."  Water and Sewage Works.

Gilderhus, P. A.  1967.  "Effects of Diquat on Bluegills and Their Food
Organisms."  The Progressive Fish-Culturist, 29:67-74.

Gillett, J. W. (ed.)  1969.  "The Biological Impact of Pesticides."  In
the Environment Proceedings of the Symposium, August 18-20, 1969 at Oregon
State University, Corvallis, OR.
      /
Guenzi, W. D. (ed.)  1974.  Pesticides in Soil Water.  Soil Science
Society of America, Inc., Madison, WI.

Henderson, C. and Q. H. Pickering.  1958.  "Toxicity of Organic Phosphorus
Insecticides to Fish."  Trans. American Fisheries Society, 87:39-51.

                                     166
-------
Henderson, C., Q. H. Pickering and E. M. Tarzwell.  1959.  "Relative
Toxicity of Ten Chlorinated Hydrocarbon Insecticides to Four Species of
Fish."  Trans. American Fisheries Society, 88:23-32.

Hermanutz, R. 0., L. H. Mueller, and K. D. Kempfert.  1973.  "Captan
Toxicity to Fathead Minnows (Pimephales promelas) Bluegills (Lepomis
macrochirus) and Brook Trout (Salvelinus fontinalis)."  J. Fisheries
Research Board Canada. 30:1811-1817.

Hughes, J. S. and T. D. Games.  1962.  "Comparative Toxicity of Bluegill
Sunfish to Granular and Liquid Herbicides."  Proceedings Sixteenth Annual
Conference Southeast Game and Fish Commissioners, pp 319-323.

Hurtig, H.  1972.  "Long-Distance Transport of Pesticides."  Environmental
Toxicology of Pesticides, Academic Press, New York, NY.

Jarvinen, A. W., M. J. Hoffman, T. W. Thorslund.  1976.  Toxicity of DDT
Food and Mater Exposure to Fathead Minnows.  EPA-600/3-76-114, U.S.
Environmental Protection Agency, Duluth, MN.

Jensen, L. D. and A. R. Gaufin.  1964.  "Long-Term Effects of Organic
Insecticides on Two Species of Stonefly Naiads."  Trans. American
Fisheries Society, 93(4):357-363.

Jensen, L. D. and A. R. Gaufin.  1966.  "Acute and Long-Term Effects of
Organic Insecticides on Two Species of Stonefly Naiads."  J. Hater Pollu-
tion Control Federation, 38(8):1273-1286.

Johnson, D. W.  1968.  "Pesticides and Fishes, A Review of Selected Liter-
ature."  Trans. American Fisheries Society, 97(4):398-424.

Katz, M.  1961.  "Acute Toxicity of Some Organic Insecticides to Three
Species of Salmonids and to the Threespine Stickleback."  Trans. American
Fisheries Society, 90(3):264-268.

Katz, M., D. L. Pederson, M. Yoshinaka and D. Sjolseth.  1969.  "Effects
of Pollution on F-ish Life."  J. Water Pollution Control Federation, 41(6).

Kemp, H. T., R. L. Little, V.  L. Holman, and R. L. Darby.  1973.  Water
Quality Criteria Data Book -Vol. 5 - Effects of Chemicals on Aquatic
Life.  U.S. Environmental Protection Agency.  NTS-PB-234 435.

Kenaga, E. E.  1974.  "2,4,5-T and Derivatives:  Toxicity and Stability in
the Aquatic Environment."  Down to Earth, 30(3):19-25.

Korn, S. and R. Earnest.  1974.  "Acute Toxicity of 20 Insecticides to
Striped Bass, Morone saxatilis."  California Fish and Game, 60:128.

Lane, C. E. and R. J. Livingston.  1970.  "Some Acute and Chronic Effects
of Dieldrin on the Sailfin Molly, Poecilis latipinna."  Trans. American
Fisheries Society. 3:489-495.
                                     167
-------
Macek, K. J. and W. A. McAllister.  1970.
Some Common Fish Family Represenatives."
Society. 1:20-27.
 "Insecticide Susceptibility of
Trans. American Fisheries
Macek, K. J., D. F. Walsh, 0. W. Hogan and D. D. Holz.  1972.  "Toxicity
of the Insecticide Dursban to Fish and Aquatic Invertebrates in Ponds."
Trans. American Fisheries Society, 3:420-427.

Macek, K. J., K. S. Buxton, S. K. Derr, J. W. Dean and S. Sauter.  1976.
Chronic Toxicity of Lindane to Selected Aquatic Invertebrates and Fishes.
EPA-600/3-76-046, U.S. Environmental Protection Agency, Duluth, MN.

Macek, K. J., K. S. Buxton, S. Sauter, S. Gnilka and J. W. Dean.  1976.
Chronic Toxicity of Atrazine to Selected Aquatic Invertebrates and Fishes.
EPA-600/3-76-047, U.S. Environmental Protection Agency, Duluth, MN.

Macek, K. J., M. A, Lindberg, S. Sauter, K. S. Buxton and P. A. Costa.
1976.  Toxicity of Four Pesticides to Water Fleas and Fathead Minnows.
EPA-600/3-76-099, U.S. Environmental Protection Agency, Duluth, MN.

Mauck, W. L., L. E. Olson and J. W. Hogan.  1977.  "Effects of Water Qual-
ity on Deactivation and Toxicity of Mexacarbate (Zectran) to Fish."  Arch.
Environmental Contamination and Toxicology, 6:385-393.

Mauck, W. L., and L. E. Olson.  1976.  "Toxicity of Natural Pyrethrins  and
Five Pyrethroids to Fish."  Archives of Environmental Contamination and
Toxicology, 4:18-29.

Mayer Jr., F. L., P. M. Mehrle Jr. and W. P. Dwyer.  1977.  Toxaphene
Chronic Toxicity to Fathead Minnows and Channel Catfish.  EPA-600/3-77-06A,
U.S. Environmental Protection Agency, Duluth, MN.

Mayer Jr., F. L., P. M. Mehrle Jr. and W. P. Dwyer.  1975.  Toxaphene
Effects on Reproduction, Growth, and Mortality of Brook Trout.
EPA-600/3-75-013, U.S. Environmental Protection Agency, Duluth, MN.

McCorkle, F. M., J. E. Chambers and J. D. Yarbrough.  1977.  "Acute Toxic-
ities of Selected Herbicides to Fingering Channel Catfish,  Ictalurea
pjjnctatus."  Bulletin of Enviromental Contamination and Toxicology,
18(3):267-270.

McKim, J. M.  1977.  "Evaluation of Tests with Early Life Stages of Fish
for Predicting Long-Term Toxicity."  J. Fisheries Research  Board of
Canada, 34(8):1148-1154.

McKim, J. M., D. A. Benoit, K. E. Biessinger, W. A. Brungs  and R.  E.
Siefert.  1975.  "Effects of Pollution on Freshwater Fish."  J. Water
Pollution Control Federation, 47(6)-.1711-1768.

Miller, M. W. and G. G. Berg  (Eds.)  1969.  Chemical Fallout Current
Research on Persistent Pesticides.  Charles C. Thomas Publishing Co.,
Springfield, 111.
                                      168
-------
Mitchell, L. E.  1966.  "Pesticides:  Properties and Prognosis."  Organic
Pesticides in the Environment.  Advances in Chemistry Series 60, American
Chemical Society, Washington, DC.

Mount, D. I. and C. E. Stephan.  1967.  "A Method for Establishing Accept-
able Toxicant Limits for Fish - Malathion and the Butoxyethanol Ester of
2,4,-D."  Trans. American Fisheries Society, 97:185-193.

Mount, D. I.  1977.  "An Assessment of Application Factors in Aquatic
Toxicology."  In Recent Advances in Fish Toxicology:  A Symposium.
EPA-600/3-77-085, U.S. Environmental Protection Agency, Corvallis, OR.

Reese, C. D. (Project Officer)  1972.  "Pesticides in the Aquatic Environ-
ment."  U.S. Environmental Protection Agency, Washington, DC.

Pickering, Q. H. and C. Henderson.  1966.  "The Acute Toxicity of Some
Pesticides to Fish."  The Ohio Journal of Science. 66(5):508-513.

Pickering, Q. H., C. Henderson and A. E. Lemke.  1962.  "The Toxicity of
Organic Phosphorus Insecticides to Different Species of Warmwater Fishes."
Trans. American Fisheries Society, 91:175-184.

Pimentel, D.  1971.  Ecological Effects of Pesticides on Non-Target
Species.  Executive Office of the President, Office of Science and Tech-
nology, Washington, D.C.

Rudd, R. L.  1964.  Pesticides and the Living Landscape.  University of
Wisconsin Press, Madison, WI.

Sanborn J. R.,  B. M. Francis and R. L. Metcalf.  1977.  The Degration of
Selected Pesticides in Soil:  A Review of the Published Literature.
EPA-600/9-77-022, U.S. Environmental Protection Agency, Cincinatti, OH.

Sanders, H. 0.   1969.  Toxicity of Pesticides to the Crustacean Gammarus
lacustris.  Bureau of Sport Fisheries and Wildlife, Tech. Paper No. 25,
U.S. Department of the Interior, U.S. Government Printing Office,
Washington, DC.

Sanders, H. 0.   1970.  "Toxicities of Some Herbicides to Six Species of
Freshwater Crustaceans."  J. Water Pollution Control Federation,
42(8):1544-1500.

Sanders, H. 0.   1972.  The Toxicities of Some Insecticides to Four Species
of Malocostrocan Crustacea.  Bureau of Sport Fisheries and Wildlife, Tech.
Paper No. 66, U. S. Government Printing Office, Washington, D.C.

Sanders, H. 0.  and 0. B. Cope.  1966.  "Toxicities of Several Pesticides
to Two Species  of Cladocerans."  Trans. American Fisheries Society.
95(2):165-169.

Sanders, H. 0.  and 0. B. Cope.  1968.  "The Relative Toxicities of Several
Pesticides to Naiads of Three Species of Stoneflies."  Limnology and
Oceanography, 13(1):112-117.
                                     169
-------
Schoettger, R. A.  1970.  Toxicology of Thiodan in Several Fish and
Aquatic Invertebrates."  Bureau of Sport Fisheries and Wildlife, Investi-
gations in Fish Control, Paper 35, U.S. Department of the Interior,
Washington, DC.

Snarski, V. M. and F. A. Puglisi.  1976.  Effects of Aroclor 1254 on Brook
Trout, Salve!inus fontinalis.  EPA-600/3-76-112, U.S. Environmental Pro-
tection Agency, Duluth, MN.

Solon, J. M. and J. H. Nair III.  1970.  "The Effect of a Sublethal Con-
centration of LAS on the Acute Toxicity of Various Phosphate Pesticides to
the Fathead Minnow, Pimephales promelas rafinesque."  Bulletin Envir. Con-
tarn, and Toxico., 5(5):408-413.

Spacie, A.  1975.  "Acute and Chronic Parathion Toxicity to Fish and
Invertebrates."  U.S. Environmental Protection Agency, Washington, DC.

Stewart, B. A., D. A. Woolhiser, W. H. Wischmeier, J. H. Caro and M. H.
Frere.  1976.  Control of Water Pollution for Cropland:  Volume I - An
Overview.  Agricultural Research Service, U.S. Department of Agriculture,
Washington, DC.

Surber, E. W. and Q. H. Pickering.  1962.  "Acute Toxicity of Endothal,
Diquat, Hyamine, Dalapon, and Silvex to Fish."  The Progessive Fish
Culturist, 24:164-171.

U.S. Department of Health, Education, and Welfare, Pesticides and Their
Relationship to Environmental Health (Parts I & II).  H.U.D. Secretary's
Commission, Washington D.C.

Walker, C. R.  1964.  "Toxicological Effects of Herbicides on the Fish
Environment."  Water and Sewage Works, 3(3).

Water Quality Criteria.  1972.  Committee on Water Quality Criteria,
EPA-R3-73-033, U.S. Environmental Protection Agency, Washington, DC.

Weed Science Society of America.  1970.  Herbicide Handbook.  Second Edi-
tion, W. F. Humphrey Press, Inc., New York, NY.

Weed Science Society of America.  1979.  Herbicide Handbook.  Fourth Edi-
tion.  W. F. Humphrey Press, Inc., New York, NY.

Wilber, C. G.  1969.  The Biological Effects of Water Pollution, Charles
C. Thomas Publishing Co., Springfield,  IL.

Wilson, D. C. and C. E. Bond.  1969.  "The Effects of the Herbicides
Diquat and Dichlobenil  (Casoron  ) on Pond Invertebrates, Part 1 - Acute
Toxicity."  Trans. American Fisheries Society, 98(3):438-443.
                                      170
-------
                                 APPENDIX C
    DISTRIBUTION COEFFICIENTS OF ORGANIC PESTICIDES IN AQUATIC ECOSYSTEMS*
 INTRODUCTION
      This report considers certain aspects of the distribution of organic
 pesticides between water and solid abiotic phases in natural aquatic (fresh
 water) ecosystems.  This study was performed for Battelle Pacific Northwest
 Laboratory in support of a larger study of the mobility and transport of the
 pesticides in natural riverine ecosystems.
      The report considers three main points:
      1.  A discussion of the molecular and environmental parameters that
          affect the distribution of the pesticides in natural systems,
      2.  A discussion of mathematical expressions useful in describing the
          distribution (or partitioning) of the pesticides between the
          aqueous and solid phases, and
      3.  The presentation of estimated distribution coefficients for the
          specific pesticides of interest in this study.
       The first two points are considered in Part 1 of this report, while the
 estimated distribution coefficients are presented in tabular form in Part 2.
*Submitted to Battelle, Pacific Northwest Laboratories, by R.N.  Dexter, URS
 Company, as final report of the services provided under consultant agree-
 ment B-62522-B-L.
                                    171
-------
                      PART 1:  ADSORPTIVE EQUILIBRIA

The discussion below is based both on a consideration of general partitioning
theory and a review of pertinent literature concerned with adsorption in
natural systems.  Many of these points considered can be found in reviews by
Hamaker and Thompson, 1972; Zettlemoyer and Micale, 1971; and Bailey and
White, 1970.

ADSORPTION

For the discussion below, equilibrium between the adsorbed and solution
phases is assumed.  At this point a balance of forces is established and
the chemical potential or activity of the sorbate (pesticide) must be
the same in the solution and on the surface of the solid matrix.  A con-
sideration of the intermolecular interactions which give rise to these
chemical potentials is then  informative  in defining the behavior to be
expected for different types of sorbates and in determining the parameters
of natural systems which can be expected to affect the chemical potentials
and thus the adsorptive equilibria.

Aqueous Solution
It is well recognized  that  liquid water  is anomalous  in  its behavior com-
pared to similar  chemical  species.  This results  primarily from the high
polarity of the water  molecules which produces  strong  internal  hydrogen
bonding.  The  dipole  also  generates  strong electrostatic  attractive forces
between water  molecules  and ionizable and polar moieties  of organic  and
 inorganic compounds.
                                     172
-------
At the same time, the hydrogen bonding among water molecules leads to con-
siderable internal structure in liquid water.  As a result, the introduction
of solute fprces the rearrangement of the normal structure of liquid water in
the vicinity of the solute molecules.  This restructuring generally requires
an energy contribution to compensate for the entropy change.  In a practical
sense, this acts as a repulsive force opposing the accommodation of the
solute.  This apparent repulsion from water can force strong associations
between certain molecules and  is referred to as hydrophobic interaction (or
hydrophobic bonding).  As a general rule, the entropic contribution and thus
the strength of the hydrophobic interaction  is a function of the effective
size .of the solute molecule in solutions (Franks, 1975).

The activity, then, is a function of both the concentration and the balance
between the attractive electrostatic forces  and the entropy-generated repul-
sion.  Solute compounds can thus he ranked based on their size and their
ability to participate in electrostatic  interactions.  For example, totally
ionized inorganic solutes are  small species  with high charge densities
leading to low  aqueous activity coefficients as reflected in their relatively
high  solubilities.  On the other end of  the  scale, non-polar hydrocarbons
interact only through relatively weak van der Waals interactions.  For these
compounds, the  hydrophobic interaction is strong resulting in high activity
coefficients and  low solubilities.  The  solubilities  and  activity coeffi-
cients of  hydrocarbons correlate well with the  size of these molecules
(Tsonopolus and  Prausnitz, 1971; Frank and Evans, 1945).

The majority of  organic pesticides  fall  somewhere between these extremes.
Most  have  some  charge localization  arising from heteroatoms  in the structure,
                                   173
-------
particularly oxygen, and some have ionizable hydrogens.  For these compounds,
the interactions with water are complex functions of both the electrostatic
and hydrophobic interactions.  Within limits, the relative activities of a
group of compounds with the same polar moiety will correlate with the size of
the non-polar substituents.  Conversely, from such correlations,, the absolute
contribution of the polar moiety can be estimated to generate empirical
additivity rules for ranking other similar compounds (Tsonopolus and Prausnitz,
1971).

These activity relationships are complicated by two factors.  First, substi-
tution of various moieties on the parent molecule can  have secondary impacts,
primarily by withdrawing or  contributing electrons to  polar sites or sites
with  active hydrogens.  For  example, monosubstituted nitrophenols as a group
are much less soluble than phenol (solubility = 93 g/1).  However, j3- nitro-
phenol (solubility  = 2  g/1)  is nearly  an order of magnitude less soluble
than  the m- or £- substituted compounds  (m-  nitrophenol,  solubility - 14 g/1;
£- nitrophenol, solubility - 17 g/1) (Morrison and Boyd,  1966,  pp. 790-792).
Second, natural waters  are not pure, but are rather complex and variable
solutions.  The principal  parameters affecting the  activity of  dissolved
species  (particularly organics) are  pH,  ionic strength and type of  ions, the
quantity  and  nature of  dissolved  (and  colloidal)  organic  matter, and tempera-
ture.

pH
The  pH  controls  the speciation of ionizable acid  and  base groups; the  ionized
forms interact more strongly with water, e.g.,  are  more soluble.
                                     174
-------
Ionic Strength
The ions of natural dissolved salts generally tend to increase the normal
ordering within the liquid such that the ability to accommodate organic
compounds in reduced ("salting out") and hydrophobic interactions are
increased.

Dissolved Organic Matter
Natural dissolved organic matter (DOM) consists principally of refractory
polyelectrolytes resulting from the degradation of biological materials
(Christman and Minear, 1971).  The DOM form stable solutions which can
scavenge and suspend pesticides either through electrostatic (ion-ion,
ion-dipole, or ligand) interactions or, for less polar materials, through
hydrophobic interactions with non-polar sites in the DOM matrix.  The net
result  is to reduce the dissolved concentration (and thus the activity) of
the pesticides in solution without necessarily decreasing the (analytically
determined) concentration.  The interactions between DOM and pesticides may
in turn be altered by changes in either pH or ion content, which will affect
the degree of ionization, the effective charge density, and the three dimen-
sional  structure of the DOM molecules.
Temperature
In all  cases, interactions are affected by temperature changes.  In general,
as the  temperature increases the activity decreases for polar species while
it may  increase for non-polar compounds.  Over the normal range of temperature
fluctuations exhibited by natural systems, the effect is small for most
compounds.
                                     175
-------
Surface Interactions
Interactions between the sorbate pesticide compounds and sites on the
absorbent solid matrix are entirely electrostatic (excluding chemisorption)
and primarily a function of the polarity of thf sorbate molecule,, ranging
from weak van der Waals to ion-ion bonding.  Natural soil and and sediment
matrices are either inorganic mineral grains, usually coated or aggregated
with organic polyelectrolytes, or detrital organic particles (Kononova,
1966).  In either case, numerous sites are available which carry weak to
strong charge localizations, usually with a net negative charge exhibited by
the whole particle (Neihof and Loeb, 1972).  To the extent that adsorption is
a surface phenomena, smaller particles will show higher mass-normalized
concentrations of pesticide due to their greater specific surface area
(Leland, et^L, 1973).

In addition, many inorganic particles in natural environments contain pores
and interstices between crystallization planes which allow sorbate molecules
to diffuse  into the interior of the particle.  (Knight  and Tomlinson, 1970.)
Similarly,  pesticides may  be capable of diffusing  into  the interior  of
detrital organic particles in  a fashion similar to  passive diffusion through
biological  membranes.   In  both cases, quantitative  differences may exist  in
the interactions between the  sorbate and the readily accessible  surfaces  of
the particles  and the  interior sites.   Further, exchange  rates with  the
 interior would  be expected to  be much slower than  with  surface sites.

 Both  pH  and ionic strength will  affect  the  characteristics of the surfaces.
The pH effectively  gives  a measure of the  H or OH" ions  available to
 satisfy  specific  acidic or basic moieties  on the  surface, while  counterions
                                     176
-------
act less specifically to satisfy residual charges sites.  Due to these
facto.s, rather complex adsorption trends may develop resulting from compet-
itive ion exchange, ligand, and dipole Interactions between surface sites,
natural ions, and polar pesticides.  Non-polar pesticides would not be
significantly affected by these changes.

One factor which must be considered, but is not often recognized, is that in
adsorption from aqueous solutions the water molecules themselves interact
with the surface and are in competition for polar sites with all other
adsorbates.  Even in cases where the relative binding strengths of water
molecules may be weak, the predominance of water molecules in natural systems
(i.e., dilute solutions) makes them important contributors to the overall
process.

The factors discussed above will affect the adsorption are summarized in
Table 1.  Based on a consideration of the dominant types of interactions
(hydrophobic or electrostatic).

                                  Table 1
     SUMMARY OF ADSORPTION INTERACTIONS IN NATURAL AQUEOUS ENVIRONMENTS
  TYPE OF
INTERACTION                   EFFECTOR VARIABLES
                                                 ENVIRONMENTAL
                   MOLECULAR           AQUEOUS PHASE	SOLID PHASE
Hydrophobic      Molecular Size        Ionic Strength,     Surface Chemistry
                                       DOM* Temp           Temp
Electorstatic    Charge Density,       Ionic Strength,     Surface Chemistry,
                 Polarizability,       pH, DOM* Temp       Ionic strenght,
                 Acid/Basic Moieties                       pH, Temp
*DOM acts to reduce the dissolved concentration.
                                     177
-------
EQUILIBRIUM RELATIONSHIPS

Theoretical Basis
To provide useful data for predicting the distribution behavior it is desire-
able to establish adsorption equilibrium relationships between the concentra-
tions in the aqueous and sorbed phases, i.e.,, adsorption isotherms.  The
primary intent is to obtain mathematical expressions, hopefully not overly
complex, for the isotherms.

A number of isotherm equations are available which have been examined as
to their applicability to  adsorption from solution in natural systems.
These include the Freundlich, Langmuir  and Brunauer-Emmett-Teller  (BET)
isotherms.  Of these only  the last two  have  any coherent theoretical basis
(Adamson,  1976).  Each is  based on different assumptions as to the processes
leading to adsorption and  each finds use in  explaining different adsorption
systems.   It should be noted that most  of the theoretical  studies  in develop-
ing  and justifying these expressions has relied on well characterized,
vapor phase systems  (Adamson, 1976).

Due  to  the strong and specific binding  arising from  ion-ion  interactions,
it would  be anticipated  that metal  salt pesticides would be  the most  likely
to follow a Langmuir type  isotherm.  Conversely,  adsorption  of non-polar
organics,  for which  hydrophobic  interactions perdominate,  would be expected
to  include multilayer formation  at  higher  concentrations and  thus  more
 likely to be represented by a  BET type isotherm.   Similarly,  BET would
be  a likely model  isotherm for organic pesticides of intermediate  polarity
                                    178
-------
since electrostatic behavior could predominate at low concentrations, but
hydrophobia interactions would undoubtedly be important as the  aqueous
solution approaches saturation.

At the same time, neither Langmuir nor  BET  isotherm  expressions  have not been
utilized to any great extent with  natural systems.   The primary  argument has
been that natural adsorbents do not present  a homogenous surface (constant
adsorption energies independent of surface  coverage) which  is required for
these equations to be applicable (Adamson,  1976).  As  a result,  the most fre-
quently encountered expression  is  the Freundlich  isotherm,  which while having
no theoretical basis, is a semi-logarithmic  type  expression with empirically
determined parameters and has been used to  fit many  observed adsorption
relationships.

However, the present body of data which has  popularized the Freundlich
isotherm is not free from criticism.   First,  the perponderance  of  studies
have been performed  by  soils chemists  and engineers primarily concerned with
the factors controlling  the  biocidal activity of agricultural pesticides.   In
the majority of these studies,  the aqueous  concentrations  of pesticides
utilized were often much higher than normally encountered  in natural  systems,
the experimental  levels  often  approaching or  even exceeding the solubilities
of the test compounds (Hamaker  and Thompson,  1972).   In some cases this was
done to maintain  detectable  levels of  the pesticides  in all phases;  in other
cases, levels approximating  field application levels  were  used.

In addition most  studies have  examined a relatively limited range  of  aqueous
concentrations compare  to what  is encountered in natural  systems.  With
                                     179
-------
this limited experimental base, it is difficult to insure that the data can
be extrapolated to either higher or lower concentrations or that the isotherm
expression is truly representative of the tdsorption behavior.

At higher concentrations, particularly as the solubility is approached,
a number of effects which can give rise to non-linear isotherms ean occur.
1)  At higher concentrations, solute-solute intermolecular interactions
may increase particularly for more hydrophobic compounds.  This results in
a decrease in the activity coefficient and thus a non-linear relationship
between the aqueous activity and the concentration such that adsorption
would be relatively diminished (other factors being invariant)j i.e, l/nl.
All of these effects  are greatly diminished at reduced concentrations when
the opportunity for solute-solute interactions either in solution or on the
surface  and the number of occupied surface sites  are all minimal.

It can be  argued  convincingly  that these conditions are most  likely  in
natural  systems  (except, of  course,  agricultural  land receiving  direct
                                    180
-------
applications), both on the basis of measured residue levels which are rou-
tinely observed at very low concentrations and by considering that the
residue undoubtedly undergo a number of adsorption-desorption steps during
mobilization from the site of application.  Under these conditions, it would
seem apparent that, since the probably sources of non-linear adsorption are
virtually eliminated, 1/n should approach unity for those systems where the
Freundlich isotherm is applicable.  Note also that at low concentrations both
the Langmuir and BET isotherms reduce to linear forms.

A further criticism of the available data is concerned with the interpreta-
tion of the kinetics of the experimental results.  Where rate information is
available, the usual behavior observed is rapid initial uptake (<24 hours)
followed by slow uptake which increases the adsorbed concentration by about
10% over a period of one to two months.  Desorption experiments performed
immediately after adsorption generally yield reversible adsorption isotherms
(at least when the aqueous concentrations of the solute are below saturation).
However, desorption and residue recovery experiments performed after long
equilibration periods have often,  but not always, shown a portion of the
adsorbed residues to be more strongly retained in the solid matrix than would
be indicated from the adsorption  isotherm (Hamaker and Thompson, 1972, pp.
92-97).

Two alternatives have been considered to explain this behavior;  1) chemisorp-
tion (Huangand Liao, 1970) and 2)  migration of the residues either to stronger
binding sites not entirely occupied during the initial adsorption, or to the
interior of the particle.  (Saha,  et al., 1969).  Chemisoi*ption seems unlikely
                                     181
-------
since the residues can be recovered in unaltered form,  and considering the
energy requirements which would be required to produce chemical binding
without destroying the pesticide molecule.  Simple migration to strongly
binding surface sites would not seem to require such long periods before
their effect would be observed.  Obviously, the simplest and most reasonable
explanation is the physical migration of the pesticide molecules into the
interior of the particle, either through pores and between the crystalliza-
tion planes of laminar clays (e.g. montmorillonites), or below the surface of
organic detrital particles.  With organically coated particles, it is possible
that the migration of polar and ionic pesticides may also yield inherently
stronger binding to the inorganic matrix (Burns and Andrus, 1970).  In most
cases, however, it appears likely that the slow rates of intraparticle
diffusion which would be sufficient to explain the slow desorption and
apparent incomplete recoveries.

Of major concern at this point  is the implication of these results in esti-
mating the mobility of the pesticide residues in natural systems.  The
major conclusions that can be drawn are 1) that the majority of experimental
K,values underestimate the actual distribution coefficient applicable
in natural systems, and 2) that many of the available desorption studies
do not realistically  represent  the behavior of the residues 1ri natural
systems.
In further support of these conclusions is the observation that most  studies
examining the migration of pesticides in  actual field situations  indicate
that the movement of  even  relatively soluble  herbicides  is generally  limited
and  much  less  than what would  be  predicted on the  basis  of laboratory
equilibrium  or  soil  leaching experiments.
                                    182
-------
Natural Systems
The information discussed above should be considered within the framework
of natural systems to discern the implication of equilibrium distribution
coefficients for predicting environmental mobility associated with the
movement of water.

The first situation is the movement of pesticides from the site of Initial
application, e.g., agricultural land.  Two major processes can be defined
1) leaching via percolation through the soil in the groundwater, and 2)
direct overland runoff associated with heavy rains or excess irrigation
water.  The great predominance of soil mass compared to water in groundwater
flow indicates that movement of even slightly adsorptive compounds should be
very slow.  On the other hand, runoff events are rapid with relatively
short contact times between the runoff water and the soils in the fields
compared to the probable rates of desorption of most compounds.  In addition
contact with the soil-incorporated residues is minimal, except for contact
with that portion of the soil which is itself mobilized by the runoff.  In
many situations, this latter soil component probably contributes the largest
fraction of mobile pesticides, independent of the strength of adsorption
of the residues.
As a result of these effects, neither the mass moved nor the distribution
between particulate and water fractions of the residues could be expected
to be strong functions of distribution coefficients.  Rather, both the
mass movement and the total concentration of the residues should correlate
with the corresponding parameters for suspended soil particles.  Such behavior
has been noted for storm-generated runoff from experimental plots (Donigian,
et__a]_L, 1977).
                                     183
-------
In permanent water bodies (rivers, lakes, etc.), however, the situation
is markedly different.  Since the residues have already undergone numerous
adsorption/desorption steps in traveling to the Mater body the aqueous
concentrations are much reduced compared to levels in agricultural usage.
Further dilution normally occurs soon after introduction into the system.
The effect should be reduced the pesticide concentrations to the range
of linear activity-concentration relationships for the residues in both
the aqueous and solid phases, and thus linear distribution coefficients, K.,
should be applicable.

Further, the solid particles are always completely hydrated.  Meiny of
the physical effects which can alter the mobility of pesticides in soils
as a result of wet/dry cycles are eliminated, e.g., collapse of pores
and voids in organic coatings on drying the slow rehydration of interior
binding sites, and the swelling of the inter-laminar spaces of clays.
It is reasonable that this stability of  the solid matrix would tend to
eliminate differences in the long term adsorption/desorption behavior
and, by keeping intraparticle voids and  pores open, facilitate exchange
between the solid  and aqueous phases.  These points would further argue
for linear, equilibrium  adsorption.
Summary
From the foregoing discussions the following conclusions can be made:

1.   The majority of the available  adsorption  data  underestimates the  actual
     strength  of  the adsorption  on  solid matrices.
2.   The majority of the  available data also oversimplifies the equilibrium
      adsorption relationships.
                                      184
-------
3.   In natural runoff from agricultural land receiving direct application of
     pesticides, the mobility of any pesticide and its relative distribution
     in the mobile aqueous and $ottd ph&ses ftrt greftoffltnantly functions of
     the physical processes taking place, with adsorption-desorption consi-
     derations of lesser importance.  It is probably not reasonable to apply
     distribution coefficients or even more complex adsorption isotherms to
     these situations.

4.   In permenant water bodies receiving indirect pesticide inputs, linear
     adsorption isotherms would be applicable, and the relative distribution
     should be adequately characterized by a single distribution coefficient,
             PART 2:  VALUES OF THE DISTRIBUTION COEFFICIENT, Kd

The estimation of reasonable K . values for the pesticides of interest in
this study is difficult for three reasons.  (1)  Fundamental physical
chemical properties, e.g., solubility, have been determined for only a
few of the compounds and even the data which have been reported are not
always reliable.  (2)  Even empirical parameters, such as the octanol-water
partition coefficient, which would be suitable for ranking the adsorptability
and estimating approximate values for K., are not readily available.
(3) The adsorption data which have been reported in the literature often
suffer from the limitations discussed in Part 1.  In addition, the data for a
single compound often vary by more than an order of magnitude, reflecting
both artifacts of different experimental techniques and real variability

                                    185
-------
resulting from differences in the adsorptive matrix, particularly when
comparisons are made between pure clays, natural loams,and natural muck or
peat soils.

The Kj values presented below were estimated for an "average" or "normal"
freshwater system.  The solid matrix is assumed to consist of silty-sand of
about l%.to 3% organic matter.  The aqueous phase is assumed to have low
total solids, pH of between 7 and 8 and to be unpolluted by large quantities
of detrital organic matter.

K. values for some non-polar  (organochlorine) pesticides and related
compounds have been predicted previously, based on fundamental physical
chemical properties (Dexter  and Pavlou, 1978).  Such an approach was not
viable for this study, however, due in  part to the  lack of reading available
fundamental data for  all  of  the compounds, and in part to the time restraints
encountered.

The  values presented  are,  of necessity, fairly rough estimates based on
careful  consideration of  the adsorption coefficients and relative  adsorption
data available  in  the literature,  and  on  a consideration of  the relative
molecular  structural  contributions  to  the adsorptive interactions, e.g.,
molecular  size, polarity  of  heteroatom moieties,  number of polar  groups,
  etc.  As  a  first  step, the  molecular  structures  were compared  and the
pesticides within  each group ranked according to  probable relative adsorp-
tion strength.  From  the  previous calculations  and  from selected  literature
references,  Kd  values for some  of the  pesticides  could  be estimated  with
 some certainty.   By further comparisons of  the  molecular  structures  of
                                      186
-------
  these "marker" compounds with the ranked pesticides, appropriate K ,
  values were estimated for natural systems.
  The values are presented in tables for each pesticide group with an approxi-
  mate ranking of the compounds within each group beginning with the highest
  Kd value.
  Each table is followed by a short discussion of the rationale for the esti-
  mated K  values and any available supporting literature data.
                          ORGANOCHLORINE PESTICIDES
A.   Aromatic                 K,
          DDT                1 X 105
          TDE (ODD)          5 X 104
          Tetradiflon        5 X 104
          Methoxychlor       1 X 104
          Kelthane           1 X 104
          Dichlone           5 X 10*
B.
Aliphatic
Aldrin
Isodrin
Chlordane
Toxphene
Mi rex
Heptachlor
Endrin
Dieldrin
Chlodecone
Endosulfan
Lindane
BHC

7 X
7 X
5 X
5 X
5 X
1 X
1 X
1 X
5 X
5 X
1 X
1 X

104
104
104
104
104
104
104
104
103
103
103
103
                                     187
-------
Rationale:  These compounds by and large are the most easily considered since
they contain relatively few polar moieties.  As a result adsorption is due
primarily to hydrophobic interaction and van der Walls forces.  For this
reason relative adsorption will be correlated with the molecular size with
corrections for polar groups and non-conjugated double bonds.  Further,
values of K . can be approximated from theoretical calculations (Dexter
and Pavlou, 1978).

Some literature support for the values can be obtained from the literature.
Values for Kd for DDT of approximately 1 X 10  have been reported in
soils by Shien et al. (1974), and Pavlou et al. (1974).  The value for
lindane was reported by Lotse et al. (1968) and Boucher and Lee (1972).
Values of distribution coefficients for some of the other compounds have
been reported, but most appear to be far from a reasonable range (Hamaker
and Thompson, 1972).
                                    188
-------
ORGANOPHOSPHATE PESTICIDES

A.   Aliphatic Deriatives               Krf
           Ethion                      5  X 102
           Disulfoton                  5  X 102
           Counter                     5  X 102
                                            ?
           Demeton                     5  X 10
           Phorate                     5  X 102
           Dioxathion                  1  X 102
           Malthion                    1  X 102
           Oxymeton methyl             1  X 102
           Dichlorvos                    50
           Ethoprop                      50
           Phosphamidon                  50
           Dicrotophos                   50
           Mevinphos                     50
           Naled                         50
           TEPP                          50
           Dimethoate                    25
           Monocrotophos                 25
           Acephate                      25
           Trichlorphon                  10
           Methamidophos                 10
B.   Phenyl Derivatives
           EPN                         1  X 103
           Carbophenothion             1  X 10
                                    189
-------
                                       Kd
          Dicapthon                   5 X 102
          Fenthion                    5 X 102
          Parathion  ethyl             5 X 102
                                           2
          Parathion,  methyl           3 X 10
          Ronnel                      2 X 102
          Stirofos                    2 X 102
          Chlorothion                2 X 102
          Famphur                    2 X 102
          Dyphonate                  2 X 102
          Ciodrin                    1 X 102
          Crufomate                  1 X 102
C.   Heterocyclic Derivatives
                                           2
          Chlorofenvinphos           5 X 10
          Phosalone                  5 X 102
          Imidan                     5 X 102
          Azinophosethyl             1 X 102
          Azinophosmethyl            1 X 102
          Diazinon                     50
          Methidathion                 50
          Chlorpyrifos                 50

Rationale:  The phosphate-based pesticides range  from  large,  complex  mole-
cules  to  relatively small.   All have polar moieties  ranging  in activity
from simply electron-rich  heteroatoms  to reasonably  strong acidic hydrogens
                                      190
-------
and basic moieties (e.g., heterocyclic nitrogens).  As a result, the pesti-
cides themselves show a wide range adsorption strength.

Limited literature references are available.  Strong binding to soils was
reported for chlorofenvinphos, carbophenothion, dioxathion and dichlofen-
thion (Inch, et al., 1972), azinophosmethyl (Helling, 1971), and for malthion
(Konrad, et al., 1969).  K. values observed for parathion include 76
(Saltzman, £t a]_._, 1972), about 120 (Bowman and Sans, 1977), and 500 (Leen-
heer and Ahlrichs, 1971).
 CARBONATE PESTICIDES
 A.    Methyl carbonates                 K.

           Bux (metalkamate)          5 X 102
           Carbonfuran                5 x 102
           Carbaryl                   5 X 102
           Propoxur                   1 X 102
           Methiocarb                 1 X 102
                                            2
           Mexacarbate                1 X 10
                                            2
           Aminocarb                  1 X 10
           Chlorpropham               1 X 102
           Propham                      50
           Benomyl                      10
           Karbutilate                   2
                                    191
-------
B.   Thiocarbamates
          Thiram                     5 X 102
          CDEC                       5 X 102
          Butyl ate                   5 X 102
                                           ^
                                           2
Vernolate                  1 X 102
          EPIC                       1 X 10'
          Molinate                     50

Rationale:  The carbonate (urethane) moiety is relatively polar and con-
tributes markedly to the greater solubility and reduced adsorption of
these compounds compared to the organochlorine pesticides.  Unfortunately,
few fundamental parameters are readily available for these compounds, nor
have extensive adsorption data been reported.

Octanol-water partition coefficients for a number of simple analogs
are all about four orders-of-magnitude less than DDT (Leo et _al_._, 1971),
indicating their low adsorption potential.  Values of K. for carbaryl
(= 125; Leenheer and Adhlrichs, 1971), benomyl (= 4.5; Austin and Briggs,
1976), and propham (= 51; Briggs, 1969) have been reported.  The relative
adsorption of some thiocarbamates has been reported by Gray and Weienrich,
(1968).
                                    192
-------
     AMINE AND ANILINE PESTICIDES

Pesticide                              Kd
     Diphenamide                     1 X 102
     Pronamide                         50
     Alachlor                          50
     Propachlor                        50
     Propanil                          10
     Carboxin                          10
     Bensultde                          5
Rationale:  The amide pesticides are chemically similar to the carbonates
and should exhibit similar  adsorption characteristics.   There are generally
smaller molecules than the  carbonates but  this  factor should be compensated
by the reduced ftumoer of polar consttutents.  The less adsorptive amides
have N-H 'groups wtiteh iShrotrtd hydrogen borrd with water molecules.
                                    193
-------
        NTIROANILINE PESTICIDES

Pesticide                              Krf
     Phenoxalin                      1 X 103
     Butralin                        5 X 102
     Profluralin                     5 X 102
     Trifluralin                     5 X 102
     Benefin                         5 X 102
     Fluchloralin                    1 X 102
     Dinitramine                       50
     Nitralin                          50

Rationale:  The nitroanilines contain relatively few non-conjugated polar
groups to contribute to electrostatic interactions.  Being quite large
molecules, hydrophobic interactions should be strong, resulting in large
KJ'S which are primarily dependent on the size of the substitution.
The ranking presented is supported by adsorption studies on soils (Harvey,
1974)   The high K, values are indicated in soil adsorption studies
by Helling (1971), Majka and Lavy (1977), and Grover (1974).  Krf values
of 1.6 X 10  and 2.8 X 10  have been reported for profluralin and
butralin, respectively (Carringer, et al., 1975); but these values were
based on adsorption on pure soil organic matter and thus are probably
higher than would  be representative of  "average" soils and sediments.
                                    194
-------
            TRIAZINE PESTICIDES

Pesticide                              Kd
     Prometome                          8
     Ametryne                           8
     Simazine                           6
     Atrazine                           5
     Cyanazine                          3
     Propazine                          1
     Metribuzin                         1
     Aminotriazole                      1

Rationale:  The triazine herbicides are all compact molecules containing
at least one polar-ionizable group, usually amino-hydrogens, available
to hydrogen bond with water.  As a result, these herbicides are quite
soluble and exhibit low adsorption on soils.

The ranking presented above has been observed in soil adsorption studies
by Helling (1971) and Rogers (1968).  Hurle and Freed (1972) reported Kd
values of 2.2. to 4.3 and 4.1 to 8.2 for atrazine and simazine, respectively,
on silt loam.  Average Kd values for adsorption by 25 soils were reported
to be:  propazine, Kd = 2.0; atrazine, Kd = 2.7; simazine, Kd = 3.7;
and prometone, Kd = 7.8 (Talbert and Fletchall, 1965).  Kd values of
approximately 4 were reported for both cyanozine and atrazine (Majka and
Lavy, 1977), while K, for atrazine have also been noted at about 2.8
(Dao and Lavy, 1978).  Liu et jiK, (1970) observed Krf values for ametryne
ranging from 2 to 10, depending primarily on the amount of organic matter
in the soils.  The latter authors reported a Kd for ametryne of approxi-
mately 150 for a muck (very high organic content) soil.
                                  195
-------
         ORGANIC ACID PESTICIDES

A.   Aliphatic Acids and Esters         K.
          Glyphosate                    0.2
          Trichloroacetic Acid          0.2
          Dalapon                       0.2

B.   Aromatic Acids and Esters
          Bifenox                     0.5 - 1
          Chloramben                  0.5 - 1
          Dicamba                     0.5 - 1
          DCPA                        0.5-1
          Fenac                       0.5 - 1
          Naptalam                    0.5-1
          Picloram                    0.5 - 1
          propargite                  0.5-1

C.   Phenoxy Compounds
          Erbon                       1 X 102
          2,4-D                         1.0
          MCPA                          1.0
          2,4,5-T                       2.0
          Silvex                        2.0
          Dinoseb                         10
          Endothall                     0.2
                                   196
-------
Rationale:  The organic acid pesticides are treated together since they
share the common feature that, with few exceptions, the activity of the
carbonic acid moiety is sufficient to make these compounds readily soluble
and to exhibit minimal adsorption.  The two exceptions to this generaliza-
tion are dinoseb, a weakly-acidic phenol, and erbon, an ester.  Both of
these latter compounds should show increased adsorption due to strong
hydrophic interactions not countered by solibilizing hydrogen or ionic
bonding.

The low adsorption of these compounds have been reported for dicamba, pic-
loram, fenac and 2,4-D (Helling,  1971); picloram (Farmer and Aochi, 1974;
Gaynor and Volk, 1976; Grover, 1971; Davidson and McDougal, 1973); 2,4,5-T
(O'Connor and Anderson, 1974); 2,4-D (Grover, 1973); picloram and 2,4-D
(Khan, 1973); picloram and 2,4,5-T (Majka and Lavy, 1977); dicamba, picloram
and 2,4-D (Grover, 1977); and dicamba  (Garringer, et al., 1975).
                                    197
-------
              UREA PESTICIDES
Pesticide                              K.
                                           2
     Chloroxuron                     5 X 10
     Chlorbromuron                   2 X 10
     Linuron                         1 X 102
     Diuron                          1 X 102
     Monuron                           50
     Fluometuron                       20
     Fenuron                           10

Rationale:  The urea pesticides are chemically similar to the carbamate and
amide pesticides and show the same range of K. values.  The molecules are
not small, but polar and hydrogen bonding moieties (N-H and C=0) decrease the
aqueous activity.

A relatively large number of studies have been reported for these compounds.
The results are summarized below as a table of observed Kd values in soils.
The references are indicated by the numbers in parentheses and are noted at
the end of the table.
                                    198
-------
Pesticide                             Kd
     Chloroxuron                      40-110(2^ 650^
     Chlorbromuron                    217^
     Linuron                          50-250(2); 10.2-15(3>; 210(4); 154(5)
(1)  Geissbuhler, et al., 1963
(2)  Hance, 1971
(3)  Hurle and Freed, 1972
(4)  Lambert, 1967
(5)  Briggs, 1969
     Diuron                           85-12; 70; 94^
     Monuron                          33(4); 29(5)
     Fluometuron                      22
     Fenuron                          15
                                     199
-------
Miscellaneous Pesticides

The miscellaneous pesticides are presented in the order they were provided
by Battelle.  The name of the compound is followed by the estimated K.
values.  Any supporting literature references are indicated by numbers which
refer to the list following the table.

Pesticide                              Kd
     Allethrin                       1 X 104
     Pyrethrum                       1 X 104
     Rotenone                        1 X 103
     CMA                               0.2
     DSMA                              0.2
     MSMA                              0.2
     Bentazon                          OJL(I)'
     Bromacil                          5(2)
     Pyrazon                          30(3,4)
     Terbacil                          50(2)
     Diquat                          5 X 103(5)
     Paraquat                        5 X 103(6,7)
     Chlorothalonil                     50
     Dichlobenil                       20  (8)
     Copper  Napthalate               5 X 102
     Fenbutalin  Oxide
     Mancozeb                            5
     Acrolein                          0.2
                                     200
-------
Pesticide                                K.
     Captan                              30
     Difol atari                           50
     Dinitrobutyl Phenol                 50
     Dodine                              0.5
     Methazole                            5
     Methomyl                             5
     Methyl Bromide                      10
     Norflura/on                         50
1.   Abernathy and Wax, 1973
2.   Rhodes, et _al^, 1970
3.   Fusi, et AL» 1976
4.   Jamet et Marie - Andree Piedallu, 1975
5.   Helling, 1971
6.   Damanakis, et al., 1970
7.   Burns, jet jfL_, 1973
8.   Furmidge and Ogersby, 1967
                                   201
-------
APPENDIX C REFERENCES
Abernathy, J.R., and L. M. Wax.  1973.  "Bentazon Mobility and Adsorption
      in Twelve Illinois Soils."  Weed Sci.  21:224-227.

Adamson, A. W.  1976.  Physical Chemistry of Surfaces.  John Wiley and Sons.
      New York.  698 pp.

Austin, D. J., and 6. G. Briggs.  1976.  "A New Extraction Method for
      Benomyl Residues  in Soil  and its Application in Movement  and Persis-
      tence Studies."   Pestic.  Sci.  7:201-210.

Bailey, G. W., and J.  L. White.  1970.  "Factors Influencing the Adsorption,
      Desorption and Movement of Pesticides  in Soils."  Res. Rev.  32:29-92.

Boucher, F.R., and A.  F. Lee.  1972.  "Adsorption of Lindane arid Dieldrin
      Pesticides on Unconsolidated Aquifer Sands."   Environ. Sci. Techno!.
      6:538-543.

Bowman, B. T., and W.  W. Sans.  1977.   "Adsorption  of  P_arathio.n, Fenitrothien,
      Methyl Parathion, Aminoparathion and Paroxon by Na  ,  Ca   „ and
      Fe    Montmorillonite  Suspensions." Soils  Sci. Soc. Amer. J.
      41:514-519.                                            ~

Briggs, G. G.  1969.   "Molecular Structure  of Herbicides and Their Sorption
      by Soils."   Nature.   223:1298.

 Burns, R.  G.  and  L.  J. Andrus. 1970.   Weed Res..  10:49-57.

 Burns, I.G.,  H. H.  Hayes,  and  M. Stacey.  1973.   "Physicochemical  Interac-
      tions of Paraquat with  Soil Organic Materials  and Model Compounds.
      2.   Adsorption  and  Desorption  Equilibria in  Aqueous Suspensions."
      Weed  Res.  13:79-90.

 Carringer, R.  D., J.  B.  Weber, and  T.  J.  Monaco.   1975.   "Adsorption-
      Desorption of Selected  Pesticides  by Organic  Matter  and Montmoril-
      lonite." ^. Agr. Food  Chem.   23:568-572.

 Christman, R.  F., and R.  A.  Minear.  1971.   "Organics in  Lakes,,"   In S.  J.
      Faust and J. V.  Hunter  (Eds.), Organic Compunds  in  Aquatice  Environ-
      ments.   Marcel  Dekker,  New York.   pp.  119-143.

 Damanakis, M., D. S.  H.  Drennan,  J. D.  Fryar,  and K.  Holly.   1970.   "The
      Adsorption and Mobility of Paraquat on Different Soils  and Soil
      Constituents."  Weed Res.  10:164-227

 Dao, T.  H., and T. H. Lavey.  1978.  "Atrazine Adsorption on Soil  as Influ-
      enced by Temperature, Moisture Content and Electrolyte Concentration."
      Weed Sci.  26:303-308.
                                    202
-------
Davidson, J. Mr., and J.  R.  McDougal.   1973.   "Experimental  and  Predicted
     Movement of Three Herbicides in  a Water  Saturated  Soil."  ^.  Environ.
     Qual.  2:428-433.

Donigan, A. S. Jr., D. C.  Beyerlein,  H. H.  Davis,  Jr.  and N.  H.  Crawford.
     1977.  "Agricultural  Runoff Management (ARM)  Model Vertion  II:   Refine-
     ment and Testing."  U.S.  Environmental Protection  Agency Report No.
     EPA-600/2-77-098, US  EPA, Environmental  Research  Laboratory,  Athens,
     Georgia.  293 pp.

Dexter, R. N. and S. P. Pavlor.  1978.  "Distribution  of Stable  Organic
     Molecules in the Marine Environment:   Physical  Chemical  Aspects.
     Chlorinated Hydrocarbons," Marine Chem., in press.

Farmer, W. J., and Y. Nochi.  1974.  "Picloram Sorption by Soils."  Soil
     Sci. Soc. Aroer. Proc.  38:418-423.

Frank, H.S., and M. W. Evans.   1945.   "Free Volume and  Entrophy  in Condensed
     Systems.  III.  Entropy in Binary Liquid Mixtures; Partial  Molal
     Entropy in Dilute Solutions; Structure and Thermodynamics in  Aqueous
     Electolytes." ^. Chem.  Phys.  13:507-532.

Franks, F. 1975.  "The Hydropholic Interaction."  In F. Franks (Ed,).
     Water:  A Comprehensive Treatise.  Vol.  4a.  Plenum Press,  New York.
     pp. 3-801.

Furmidge, C. C., and J. M. Osgerby.  1967.   £. Sci.  Food Agr.  18:269-173.

Fusi, P., S. Ceconni, M. Franci, and  C. Vazzana.  1976.  "Adsorption et
     Desorption de la Pyrazone par Quelques Collorides Organiques  et
     Nunreaux."  Weed Res.  16:101-109.

Gaynor, J.D., and V. V. Volk.   1976.   "Surfactant Effects on Picloram Adsorp-
     tion by Soils."  Weed Sci. . 24:549-552.

Geisshihler, H., C. Haselbach, and H.  Aebi.  1963.  "The Fate of N-(4-
     Chlorophenoxy)-Phenyl-N,N-Dimethylurea (C-1983) in Soils and  Plants."
     Weed Res.  3:140-153

Gray, R.A., and A. J. Weirich.  1968.   "Leaching of Five Thiocarbanate
     Herbicides in Soils."  Weed Sci.   16-77-79.

Grover, R.  1973.  "The Adsorptive Behavior of Acid and Ester Forms of
     2,4-D on Soils."  Meed Res.  13:51-58.

Grover, R.  1971.  "Adsorption of Picioram by Soil Colloids and  Various
     Other Adsorbents."  Weed Sci.  19:417-418.

Grover, R.  1977.  "Mobility of Dicamba, Picioram and 2,4-D in Soil  Columns."
     Weed Sci.  25:259-162.
                                    203
-------
Grover, R.  1974.  "Adsorption and Desorption of Trifluralin, Triallate and
     Diallate by Various Adsorbents."  Need Sci.  22:405-408.

Hamaker, 0. W., and J. M. Thompson.  1972.  "Adsorption".  In C. A.  Goring
     and J.W. Hamaker (Eds.). Organic Chemicals' in the Soil Environment,
     Vol. 1. Marcel Dekker, New York.pp.49-143.

Hance, J.C.  1971.  "Complex Formation as an Adsorption Mechanism for
     Linuron and Atrazine."  Heed Res.  11:106-110.

Harvey, R.  1974.  "Soil Adsorption and Volatility of Dinitroaniline Herbi-
     cides."  Weed Sci., 22:120-134.

Helling, C. S.  1971.  "Pesticide Mobility in Soils III.  Influence of
     Soil Properties."  S_. Sci. Soc. Amer. Proc.  35:743-748.

Huang, J. C., and C. S. Liao.  1970.  "Adsorptives of Pesticides by Clay
     Minerals," J. San. Eng. Div., Proc. Amer. Soc. Civil Eng., 96:SA5,
     1057-1064.

Hurle, K. B., and V. H. Freed.  1972.  "Effects of Electrolytes on the
     Solubility of Some 1,3,5-Triazines and Substituted Ureas  and Their
     Adsorption on Soils."  Weed Res.  12:1-10.

Inch, T. D., R. V. Levy, and D. Utley.  1972.  "The Mobility of Some Organo-
     phsphorus Sheep Dip Insecticides in Soil."  Pestic. Sci.  3-242-253.

Jamet Et Marie-Andree Piedallu, P.  1975.  "Etude de  Tadsorption et de la
     Desorption de la Pyrozone (Amino - 5 Chloro - 4  Pheny (-2,2H Pyrida-
     zinone-3) par Differents Types de Sols."  Weed Res.  15:113-121.

Khan, S.U.   1973.  "Interaction of Humic Acid with Chlorinated Phenoxyacetic
      and  Benzoic  Acids."   Envir. Letters.  4:141-148.

Knight,  B.  A. G.  and T.  E. Tomlinson.  1970.  In Sorption  and  Transport
     Processes in Soils, Soc. Chem.  Ind., London, p.  54-77.

Kononova,  M. M.   1966.   Soil Organic  Matter  (translated  by T.  Z. Nowakowski
      and  A.C.D. Newman,  Pergamon Press, New  York).

Konrad,  J., G. A.  Chesters,  and D.  E. Armstrong.  1969.  "Soil  Degradation
     of  Malthion, a  Phosphorodithieate  Insecticide."   Soil Sci. Soc. Amer.
      Proc.   33:259-262.

Lambert,  S* M.   1967.   "Functional  Relationships Between Adsorption  in  Soil  and
      Chemical  Structure."  ^3.  Agr.  Food  Chem.  15:572-575.

Leenheer,  J.  A.  and  0.  L.  Ahlricks.   1971.   "A  Kinetic and Equilibrium  Study
      of the Adsorption  of  Carbaryl  and  Parathion Upon Soil Organic Matter
      Surfaces."   Soil  Sci. Soc. Am.  Proc.  35:700-705.


                                    204
-------
Leland, H. V., W. N. Bruce, and N. F. Shimp.  1973.  "Chlorinated Hydrocarbons
     Insecticides in Sediments of Southern Lake Michigan," Environ. Sci.
     Techno!., 9:883-838.

Leo, A., C. Hansch, and D. Elkins.  1971.  "Partition Coefficients and Their
     Uses."  Chem. Rev.  71:525-616.

Liu, L.C., H. Cibes-Viode, and F.K.S. Koo.  1970.  Adsoption of Dimetryne
     and Diron by Soils."  Weed Sci., 13:470-474.

Lotse, E. G., D. A. Graetz, G. Chesters, G. F. Lee, and L. W. Newland.
     1968.  "Lindane Adsorption by Lake Sediments."  Environ. Sci. Techno!.
     2:353-357.

Majka, J.T., and T. L. Lavy.  1977.  "Adsorption, Mobility and Degradation
     of Cyanazine and Diuron in Soils."  Weed Sci.  25:401-406.

Morrison, R. T., and R. N. Boyd.  1966.  Organic Chemistry.  Allyn and
     Bacon, Boston.  1204 pp.

Neihof, R. A. and G. I. Loeb.  1972.  "The Surface Charge of Particulate
     Matter in Seawater."  Limnol. Oceanogr.  17:7-16.

O'Conner, G. A. and J. U. Anderson.  1974.  Soil Factors Affecting the
     Adsorption of 2,4,5-%.  Soil Sci. Soc. Amer. Proc.  38:433-436.

Pavlou, S. P., R. N. Dexter, and 0. R. Clayton.  1974.  "Chlorinated Hydrocar-
     bons in Coastal Marine Ecosystems."  In Proceedings of the International
     Conference on Transport of Persistent Chemicals in Aquatic Ecosystems,
     May 1-3, 1974, Ottawa, Canada,  pp. II-31-II-35.

Rogers, E. H.  1968.  "Leaching of Seven S-Triazgines."  Weed Sci.  26:117-121

Rhodes, R.C., I. J. Belasc, and H. L. Pease.  1970.  J. Agr. Food Chem.
     8:524-529.                '                           	

Saha, J. G., B. Bharati, Y. W. Lee and R. L. Randall.  1969.  J. Agr. Food
     Chem., 17:877-893.

Saltzman, S., L. Kliger, and B. Yaron.  1972.  "Adsorption-Desorption of
     Parathion  as Affected by Soil Organic Matter."  J. Agr. Food Chem.
     20:1223-1226.                                   ~

Shin, Y. 0., J. J. Chordan, and A. R. Wolcott.  1970.  "Adsorption of DDT
     by Soil Fractions and Biological Materials."  J. Agr. Food Chem.
     18:1129-1133.                                 ~

Talbert, R. E., and 0. H. Fletchall.  1965.  "The adsorption of Some S-
     Triazgines in Soils."  Weed Sci.  13.--46-49.
 ftU.S GOVERNMENT PRINTING OFFICE 1982-559-092/3401
                                    205
-------