EXPOSURE ASSESSMENT MODELING
           FOR
    ALDICARB IN FLORIDA
                Anderson-Nichols/
                    Engineers • Environmental Consultants • Planners

-------
                                   October 1984
   REVIEW  DRAFT
     EXPOSURE ASSESSMENT MODELING
                 FOR
          ALDICARB IN FLORIDA
                 by
               J.D. Dean
              D.F. Atwood
     Anderson-Nichols & Co.,  Inc.
         Palo Alto, CA  94303
             Final Report

        Contract No. 68-03-3116
        Work Assignment No.  23
            Project Officer

           Mr.  Lee A. Mulkey
  U.S. Environmental Protection Agency
    Environmental  Research Laboratory
           Athens, GA  30613
ATHENS  ENVIRONMENTAL RESEARCH LABORATORY
   OFFICE OF RESEARCH AND DEVELOPMENT
  U.S.  ENVIRONMENTAL PROTECTION AGENCY
           ATHENS, GA  30613

-------
                           ABSTRACT
A   modeling   study   was   performed   to   assess   aldicarb
concentrations  in  drinking  water wells in  the  vicinity  of
citrus  groves  in the state of Florida.  Areas in  the  citrus
growing   region   were   identified,  with  respect   to   the
unsaturated  and  saturated zones, in which fate and  transport
of   aldicarb  was  thought  to  be  uniquely  different.    In
addition,  an  extensive  literature search  was  conducted  to
determine  degradation  rates and adsorption  coefficients  for
aldicarb.   These  regional  and  chemical data  were  used  to
define  various  simulation scenarios.  The fate and  migration
of  aldicarb was then simulated for the unsaturated zone  using
PRZM  (the  Pesticide Root Zone Model) and the  saturated  zone
using CFEST (Combined Fluid-Energy-Solute Transport model).

Results  of  the  unsaturated zone modeling showed  that  there
were  three  statistically  distinct scenarios with  regard  to
pesticide   leaching;  "ridge"  soils  with  thick  unsaturated
zones,   "ridge"   soils  with  thin  unsaturated  zones,   and
"flatwoods"  soils.  The highest  loads leached to ground  water
from  the treated band were approximately 1 kg/ha, occurring in
areas of "ridge" soils with thin  unsaturated zones.

Combined   results  of  the  unsaturated  and  saturated   zone
modeling   showed   that,  in  general,  drinking  water   well
concentrations   should   be   low.   The   highest   simulated
concentrations  were  approximately  at the detection  limit  of

                              ii

-------
the  chemical  (5  to  6  ppb).   Highest  concentrations  were
simulated  in  the surficial unconfined aquifer system, not  in
the  Floridan,  overlain  with  "ridge"  soils  having  a  thin
unsaturated  zone.   Highest  simulated concentrations  in  the
Floridan Aquifer were less than 1 ppb.

The  effects  of well distance from the source area  were  also
investigated.   In  the surficial aquifer,  with  hydrogeologic
properties  most conducive to aldicarb transport, a well at 300
m  (1000 ft) versus a well at 91 m (300 ft) should have from  2
to  100  times  less aldicarb, depending upon  pesticide  decay
characteristics.   In  the Floridan Aquifer with  hydrogeologic
properties  most conducive to transport, a well at 300 m  (1000
ft)  should  have  from  nearly  the  same  to  10  times  less
aldicarb, again, depending upon pesticide degradation rates.

Because   of  the  regional  scope  of  the  study  and   model
limitations,  "catastrophic" situations such as the presence of
sink  holes  or leaky wells which would result in  much  higher
concentrations   were  not  simulated.   These  situations  are
discussed in the report.
                              111

-------
                           CONTENTS

                                                        Page
Abstract	   ii
Figures	   vi
Tables	 xiii
Acknowledgments	xviii

1.  Executive Summary	    1
    1.1  Purpose of Study	    1
    1. 2  Technical Approach	    2
    1.3  Environmental and Pesticide Characteristics...    5
    1.4  Summary of Results	   23
    1.5  Combined Results of Unsaturated and Saturated
         Zone Modeling	   33
    1.6  Conclusions and Recommendations	   38
2.  Florida Citrus Growing Environment	   47
    2.1  Surface and Unsaturated Zone	   47
    2. 2  Saturated Zone	   91
3 .  Chemical Fate and Transport	   124
    3.1  Aldicarb Fate	   124
    3.2  Transport Considerations	   130
    3.3  Analysis of Pesticide Model Parameters	   134
4.  Model Application and Results	   153
    4.1  Unsaturated Zone Modeling	   154
    4.2  Saturated Zone Modeling	   197
    4.3  Combined Results of Unsaturated and
         Saturated Zone	   254
References	   264
Appendices
    A  PRZM Modifications	   275
                             iv

-------
B  Plots of Soil Order Physical Properties from
   Soil Characterization Analysis	  291
C  Derivation  of First-Order Hydrolysis Rate
   Equation for Saturated Zone	  307
D  Aldicarb TTR Data from Oviedo and Davenport
   Sites for 1984	  312
E  Cumulative Frequency Distributions of Relative
   Weil-Water Concentrations	  322

-------
                            FIGURES
                                                        Page
1.1   Florida counties ranked by citrus acreage	   6

1.2   Subareas delineated for unsaturated zone modeling.   9

1.3   Matrix showing unsaturated zone modeling
      scenarios simulated	  11

1.4   Principal aquifer geometries in the citrus
      growing area	  12

1.5   Matrix showing saturated zone scenarios simulated.  18

1.6   Schematic of aldicarb environmental chemical
      pathways	  20

1.7   Frequency of annual quantity of pesticide leached
      to the saturated zone from thick Entisols and
      Ultisols	  25

1.8   Frequency of annual quantity of pesticide leached
      to the unsaturated zone from thin Entisols and
      Ultisols	  26

1.9   Frequency of annual quantity of pesticide leached
      to the saturated zone from Spodosols and Alfisols.   27

                              vi

-------
1.10  Overlap of unsaturated and saturated model areas..    34

2.1   Florida counties ranked by citrus acreage	    48

2 . 2   Annual precipitation in inches	    51

2. 3   Annual lake evaporation in inches	    52

2.4   Mean annual number of days having a minimum
      temperature of 32 F (0 C) or below	    55

2.5   Location and extent of soil orders on which
      citrus is grown	    56

2.6   Mean values of field capacity water content
      vs. depth from soil characterization analysis	    60

2.7   Mean values for saturated hydraulic conductivity
      versus depth from soil characterization analysis..    61

2.8   Mean values of soil pH versus depth from the
      soil characterization analysis	    63

2.9   Mean values of organic carbon versus depth from
      soil characterization analysis	    64

2.10  Typical configuration for bedded citrus in the
      Flatwoods Area	    71

2.11  Generalized topography of the Central Peninsula,
      Florida	    76

2.12  Typical cross-section through a ridge citrus area
      showing relative thickness of unsaturated zone....    77

                             vii

-------
2.13  Subareas delineated for unsaturated zone modeling.   79

2.14  Selected meteorologic stations for unsaturated
      zone modeling	   85

2.15  Fall of fre.e water level during one furrow
      irrigation drying cycle in single bedded
      'Ruby Red1 grapefruit groves planted on Felda and
      Immokalee soil types in the Indian River area	   90

2.16  Principal areas where the piezometric surface of
      the Floridan Aquifer rises above the water table..   97

2.17  Areas of the Floridan Aquifer which do not meet
      drinking water standards	   99

2.18  Principal aquifer geometries in the citrus
      growing area	  102

2.19  Potentiometric surface of the Floridan Aquifer....  115

3.1   Schematic of aldicarb environmental chemical              >.
      pathways	  126

4.1   Predicted and observed Aldicarb TTR in the upper
      300 cm of the soil at the Lake Hamilton site,            /
      1984	  156

4.2   Predicted movement of Aldicarb residues at the            /
      Lake Hamilton location	  157

4.3   Location of Aldicarb TTR in the soil profile              /
      > 5 ppb at the Lake Hamilton site	  159
                             VI11

-------
4.4   Predicted and observed Aldicarb TTR in the upper          ,
      150 cm of the soil at the Oviedo site	  160
4.5   Location of Aldicarb TTR in the soil profile
      > 5 ppb Oviedo	  161

4.6   Comparison of Aldicarb TTR degradation rates
      for 1984 Ovieda and Davenport data	  166

4.7   Comparison of measured and simulated Aldicarb
      TTR concentrations for two 1984 sampling dates
      at Oviedo, Florida	  168

4.8   Comparison of observed and simulated Aldicarb
      TTR concentrations for two 1984 sampling dates
      at Davenport, Florida	  169

4.9   Schematic of a unit block of citrus (one tree,
      not to scale)	.	  172

4.10  Fate of Aldicarb TTR after application to soil....  180

4.11  Geometric mean annual quantity of pesticide
      leached to the saturated zone from the treated
      band	  183

4.12  One way analysis of variance of the effect of
      irrigation on method annual pesticide mass
      leached to the saturatd zone (kg/ha)	  185

4.13  Frequency of annual quantity of pesticide leached
      to the saturated zone from Entisol scenarios	  186

4.14  Frequency of annual quantity of pesticide leached
      to the saturated zone from Ultisol scenarios	  187

                              ix

-------
4.15  Frequency of annual quantity of pesticide
      leached to the saturated zone from the Spodosol
      scenarios	  188

4.16  Frequency of annual quantity of pesticide leached
      to the saturated zone from the Alfisol scenarios..  189

4.17  Frequency of annual quantity of pesticide
      leached to the saturated zone from thick Entisols
      and Ultisols	  191

4.18  Frequency of annual quantity of pesticide leached
      to the unsaturated zone from thin Entisols and
      Ultisols	  192

4.19  Frequency of annual quantity of pesticide
      leached to the saturated zone from Spodosols and
      Alf isols	  193

4.20  Cumulative mass curve of pesticide leaching within
      the year for three representative scenarios	  198

4.21  Plan view of a hypothetical citrus grove and
      well configuration	  202

4.22  Schematic graph of pesticide concentration through
      time based on residence time under source area....  210

4.23  Schematic diagram of pesticide accumulation
      process	  211

4.24  Concentration profile for eight month residence
      time based on plug flow	  213
                               x

-------
4.25  Effect of dispersion on concentration in the
      well over time	  214

4.26  Concentration histories at variable field
      widths for the Floridan worst case scenario
      with no decay	  216

4.27  Concentration histories at variable field widths
      for the Floridan worst case scenario with decay...  217

4.28  Relationship of field width to peak concentration
      at the well for the Floridan worst case	  218

4.29  Concentration histories at variable field
      widths for the surficial worst case scenario
      without decay	  219

4.30  Concentration histories at variable field widths
      for the surficial worst case scenario with decay..  220

4.31  Concentration versus time for Floridan worst
      case with a shallow well	  224

4.32  Concentration versus time for the Floridan worst
      case with a deep well	  225

4.33  Concentration versus time for the Floridan
      average case with a shallow well	  227

4.34  Detail of Floridan average case with a shallow
      well and decay	  228

4.35  Concentration versus time for the Floridan
      average case with a deep well	  229
                              XI

-------
4.36  Detail of Floridan average case with a deep
      well and decay	  230

4.37  Concentration versus time for the surficial
      worst case	  233

4.38  Detail of the surficial worst case with the well
      1000 ft from the source with decay	  234

4.39  Concentration versus time for the two-aquifer
      worst case with the well 300 ft from the source...  236

4.40  Comparison of the two extreme hydraulic
      potential distributions	  243

4.41  Effect of Aldicarb source area surrounding
      the well	  244

4.42  20-year simulation of average Floridan case	  246

4.43  Relationship of model time step to peak
      concentrations for the Floridan and surficial
      worst cases	  249

4.44  Unsaturated and saturated zone groupings	  255

4.45  Overlap of unsaturated and saturated model areas..  256
                              Xll

-------
                            TABLES
                                                        Page
1.1   Approximate Citrus Acreage Associated With
      Each Aquifer Geometry	  14

1.2   Cases Considered for Saturated Zone Modeling	  17

1.3   Summary of Pesticide Loadings per Unit Citrus
      Block Area for the Three Final Unsaturated Zone
      Scenarios	  28

1.4   Relative Concentration Values for Indicated
      Saturated Scenarios at Three Exceedance
      Probabilities	  31

1.5   Highest Calculated Aldicarb Concentrations
      (in ppb) in the Given Combined Unsaturated/
      Saturated Catagories	  36

2.1   Summary of Bearing and Nonbearing Citrus
      (Oranges and Grapefruit) by County as of
      January 1, 1982	  50

2.2   Mean Monthly Rainfall and Evapotranspiration at
      Two Citrus Growing Locations in Florida (cm)	  53
                             Xlll

-------
2.3   Soil Samples Selected for Soil Characterization
      Analysis	  58

2.4   Citrus Irrigation Systems with Acreages for Each
      of Three Areas in the State	  67

2.5   Summary of Management Characteristics of
      Irrigation Systems	  68

2.6   Characteristics of Areas Selected for Unsaturated
      Zone Modeling	  80

2. 7   Unsaturated Zone Modeling Scenarios	  83

2.8   Hydraulic Characteristic Data for Florida Soils
      by Horizon	  87

2.9   Physical Factors Influencing Pesticide Movement
      in the Saturated Zone	  93

2.10  Generalized Stratigraphic Units and Associated
      Hydrogeologic Properties	  94

2.11  Approximate Citrus Acreage Associated With Each
      Aquifer Geometry	 104

2.12  Cases Considered for Saturated Zone Modeling	 108

2.13  Average and Worst Case Hydrogeologic Parameters
      for the Three Aquifer Configurations	 Ill

2.14  Well Rates and Depths for Saturated  Zone Modeling. 123
                             xiv

-------
3.1   First Order Rate Constants for Oxidation and
      Hydrolysis of Aldicarb in Soil	 128

3.2   First Order Hydrolysis Rate Constants for
      Aldicarb, Aldicarb Sulfoxide and Aldicarb Sulfone..l29

3.3   Koc for Aldicarb and its Daughter Products	 132

3.4   Multiple Linear Regression Coefficients for
      Aldicarb Transformation and Degradation Rates	 136

3.5   Characteristic Data for Florida Soils Used to
      Estimate Pesticide Parameters by Horizon	 138

3.6   Transformation and Degradation Rates (k) and
      Adsorption Partition Coefficient (K) Used in
      Modeling Aldicarb Fate and Transport in Florida
      Soils	 139

3.7   Comparison of Degradation Rates of Aldicarb
      Residues Estimated by Hydrolysis With Degradation
      Rates Measured in Soil and Water Degradation
      Studies	 144

3.8   Hydrolysis Rate Data Used to Calculate Activation
      Energy  Parameters	  147

3.9   Distilled Water Hydrolysis Rates of Aldicarb
      Sulfoxide and Aldicarb Sulfone	  148

3.10  Estimated Values  of Activation Energies and
      Pre-Exponential Factors for Aldicarb, its Sulfoxide
      and Sulfone	  149
                              xv

-------
3.11  First-Order Decay Rates in Florida's Ground Water
      Based on Activation Energy Analysis	  151

4.1   Aldicarb Residues (TTR) in the Soil at Davenport,
      Florida, 1984	  163

4.2   Aldicarb Residues (TTR) in the Soil at Oviedo,
      Florida, 1984	  165

4.3   Some Water Balance Components for Twenty-Four
      Unsaturated Zone Scenarios (in centimeters)	  176

4.4   Fate of Aldicarb TTR after Application to Soil....  182

4.5   Summary of Pesticide Loadings per Unit Citrus
      Block Area for the three Final Unsaturated Zone
      Scenarios	  194

4.6   Mean Percentage of Aldicarb, Aldicarb Sulfoxide
      and Aldicarb Sulfone in the Simulated Leached
      Pesticide Load	  196

4.7   Model Input Parameters for the Six Aquifer
      Types Simulated	  201

4.8   Initial Input Concentrations Used in CFEST for
      Each Aquifer System Simulated  (in ppb)	  205

4.9   Time Required to Leach 90% of the Pesticide from
      the Unsaturated Zone	  207

4.10  Summary of Results for Floridan Worst Case
      Simulations	  223

4.11  Summary of Results for Floridan Average Case	  226

                             xvi

-------
4.12  Summary of Results for the Surficial Worst Case...  232

4.13  Summary of Results for Two-Aquifer System,
      worst case	  235

4.14  Comparison of Peak Concentrations Determined by
      a Simple Plug Flow Model and by Computer
      Simulation	  239

4.15  Results of Average Versus Actual Aldicarb Mass
      Loading Simulations	  253

4.16  Well Water Concentrations (ppb) for Combined
      Unsaturated and Saturated Scenarios at 50 and
      10 Percent Exceedance Probabilities of Each	  259

4.17  Highest Calculated Aldicarb Concentrations
      (in ppb) in the Given Combined Unsaturated/
      Saturated Catagories	  261
                              xvn

-------
                       ACKNOWLEDGEMENTS
There  are  many,  many individuals who spent a great  deal  of
time  and  effort/ offering data and advise, both freely  given
and eagerly accepted.

First  of  all,  the  authors would  like  to  acknowledge  the
support  throughout  the project of Mr. Lee Mulkey of the  U.S.
Environmental Protection Agency in Athens, GA.

Dr.  Jim  Davidson of the University of Florida is thanked  for
helping  us to tap the rich resources of data and personnel  at
the  University  of Florida.  Dr. Daniel Spangler,  Jim  Jensen
and  Al  Quarles  of  the Geology Department  were  helpful  in
delineating  saturated zone scenarios and helping us find  data
for   aquifer   properties.    Mr.  Dalton  Harrison   of   the
Agricultural  Engineering Department and Dr. Lawrence  Parsons,
Dr.  Carl  Anderson, Dr. Robert Koo and Dr. Harry Woods of  the
Lake  Alfred Citrus Research Center are especially thanked  for
advice   and  data  on  soils  and  irrigation  practices   and
hospitality in arranging field trips into groves.

Drs.  R.H.  Biggs and P.G. Webb of the Fruit Crops  Department,
University  of Florida are thanked for the information supplied
on aldicarb plant uptake.

We  gratefully  acknowledge  the aid of Drs. Lamar  Miller  and
Jeff  Foran  (University of Florida) in supplying 1984  residue
                             xviii

-------
data from the Oviedo and Davenport sites.

Dr.  Johan  Smelt  of  the Institute  for  Pesticide  Research,
Wageningen,   the  Netherlands  and  Dr.  Richard  Bromilow  of
Rothampsted  Experimental  Station, U.K. are thanked for  their
discussions    on   ferrous   iron   catalysis   of    aldicarb
transformation reactions.

Mr.  Charles  Tibbals  of the U.S.  Geological  Survey  Orlando
office,  Dr.  Tony  Irwin and Dr. John Vechiolli  of  the  U.S.
Geological  Survey,  Tallahassee office and Mr. Jim  Frazee  of
St.John's  Water Management District were most helpful in their
advice  and  feedback on the ground-water modeling  parameters.
Dr.  Ann  Lemley of the Department of Design and  Environmental
Analysis  at Cornell University is thanked for her  information
on aldicarb hydrolysis rates and activation energy.

Mr.  Fredrick Bond and Ms. Chris Smith of Battelle's Office  of
Hazardous  Waste  Management provided expertise with the  CFEST
model used for saturated zone modeling.

Review  of the report was provided by Mr. Anthony S.  Donigian,
Jr.,  of  Anderson-Nichols.   Ms. Tomi Hutchins  and  Ms.  Lisa
Jowise  prepared the graphics.  Word processing was provided by
Ms.  Carol  McCullough and Mrs. Dorothy  Inahara.   Editing  and
proofing was provided by Ms. Susan Reutter-Harrah.
                              xix

-------
                           SECTION 1
                       EXECUTIVE SUMMARY
1.1  PURPOSE OF STUDY
The  pesticide Temik (aldicarb) is used extensively in  Florida
citrus  for the control of nematodes and other pests.  It is  a
systemic  insecticide  that  is highly mobile in soils  and  is
also  very  toxic.  Evidence of Temik contamination  in  ground
water  in  the  state  of Florida sparked  public  concern  and
prompted  a March 1983 suspension of the use of the compound in
all  but three Florida counties.  In September 1983, the use of
Temik  was  reinstated with the restrictions that it would  not
be  used within 90 m (300 ft) of drinking water wells and  that
application would be at half the label rate.

Because  of the lack of adequate data, it is difficult to  make
a  meaningful assessment of whether current concentrations  are
high  enough to support a general ban or to evaluate management
options  if  continued  use is permitted.  A  coordinated  data
collection  program,  modeling  study, and risk  assessment  is
being  conducted by the EPA Office of Research and  Development
to  establish current contamination levels and the health  risk
associated with the continued use of Temik on citrus.

-------
The  purpose of the modeling portion of the work is to  provide
an  assessment of the migration and fate of aldicarb in topical
Florida  citrus applications with emphasis on the potential for
leaching  and ground-water contamination.  This information can
be  used subsequently to perform exposure and risk  assessments
and  to  evaluate  the effects of management  alternatives  and
current use restrictions on Temik.

This report describes the modeling portion of the study.
1.2  TECHNICAL APPROACH
The  concentration  of a pesticide appearing in a well  is  the
result  of the interaction of two sets of factors; those  which
affect  the  supply of the compound and those which affect  the
transport  of  the  compound  from the  location  where  it  is
applied  to  the  wells in which it may  appear.   The  factors
which  affect  the  supply and transport break  down  into  two
major  groups;  those  which are compound  specific  and  those
which  are  site  specific.  The fundamental approach  to  this
problem   is  to  collect  and  analyze  information  on  those
compound-  and  site-specific factors, and then use  models  to
provide  the interactive linkages which result in estimation of
well-water concentrations.

The   first  task  using  the  above  approach  is  to   gather
information  on specific chemical properties and  site-specific
information  about  the  area of interest.  From  the  chemical
information   values  of  constants  used  in  modeling  (e.g.,
adsorption   coefficients,   decay  and   transformation   rate
coefficients)  are  determined.  The site-specific  information
is  used  to delineate a number of scenarios among which it  is

-------
felt  the  fate  and transport of  the  chemical  significantly
differ.   These scenarios may differ, for instance, because  of
climatic,  soils  or hydrogeologic variables.  The  information
gathered  is  also  used  at  later  stages  to  develop  model
parameter sets for each scenario.

A  model  or  models are then selected which can  describe  the
fate   and   transport   of  the  chemical   in   the   region.
Modifications   to   models,  which  are  made   necessary   by
idiosyncrasies  of the compound or application sites, are  made
at  this point.  If possible, models and modifications made  to
models  should  be tested and, if necessary, calibrated in  the
area of interest.

For  each  scenario,  models are run to  provide  estimates  of
well-water  concentrations  at various probability  levels.   A
probabilistic  framework is necessary because of the stochastic
impact  of  climate on transport in these systems.   Impact  of
uncertainty  in  other variables (e.g., pesticide  decay  rate,
aquifer   permeabilities)  is  assessed  by  perturbing   these
variables  in model simulations and observing subsequent impact
on model output.

In  this case, we are specifically concerned with the  compound
aldicarb  as  used  in  the citrus  growing  area  of  Florida.
Site-specific  data was gathered on the Florida citrus  growing
region.   This  information  gathering  was  divided  into  two
topical   areas;  characteristics  of  the   land  surface   and
unsaturated  zone,  and characteristics of the saturated  zone.
Factors  of  particular  interest in the unsaturated  zone  are
rainfall    depths   across   the   region,   soil    hydraulic
conductivities  and  pH,  and irrigation and  other  management
practices.   In  the  saturated zone, recharge  rates,  aquifer
hydraulic   conductivity,   water  table   gradients,   aquifer
configuration,  ground-water temperature and pH are of  primary

-------
interest.    This  information  was  used  to  delineate   both
unsaturated  and  saturated  zone  modeling  scenarios.   These
delineations   were   based  on  observed  variations  in   the
important  factors which we felt would result in  significantly
different fate and transport  characteristics.

The  next  step was to summarize chemical properties  for  this
compound  which  relate to fate and transport.   For  aldicarb,
properties  of  particular  interest  are  transformation  rate
coefficients  for the parent compound to aldicarb sulfoxide and
aldicarb  sulfone,  and  hydroloysis rates and  soil  partition
coefficients for each species.

PRZM  (the Pesticide Root Zone Model, Carsel et al., 1984)  and
CFEST  (Combined  Fluid-Energy-Solute Transport, Gupta et  al.,
1982)  were  chosen to represent the unsaturated and  saturated
zones  respectively.   Modifications were necessary to PRZM  to
accomodate   the  transformation  of  aldicarb  to  two   toxic
daughter  products and lateral drainage induced by the practice
of  bedding  citrus in some areas.  No modifications  to  CFEST
were  required.   PRZM  was verified using local  soil  residue
data  to  insure  that it adequately represented the  fate  and
transport  of  aldicarb  in the unsaturated zone.  It  was  not
feasible  to  test CFEST in this manner for the saturated  zone
applications due to a lack of data.

For  each unsaturated zone scenario, PRZM was run to produce   a
time  series and frequency distribution of pesticide  loadings.
These  loadings  were  subsequently used by  CFEST  to  produce
well-water   concentrations.    Within  each   saturated   zone
scenario,  sensitivity to factors such as distances from source
area  to  wells,  size of source area,  pesticide  loading  and
decay rates was determined.

-------
1.3  ENVIRONMENTAL AND PESTICIDE CHARACTERISTICS
This  section discusses briefly the review of literature on the
pesticide  aldicarb and the Florida citrus growing  environment
used  to develop modeling scenarios.  A more detailed treatment
of  this  material  is contained in Sections 2 and  3  of  this
report.
1.3.1  Characteristics of the Citrus Growing Environment
1.3.1.1  Unsaturated Zone—

Citrus  is  grown  throughout  most  of  central  and  southern
Florida.   The  map  of Figure 1.1 shows the  counties  of  the
state  in which citrus is grown ranked according to the acreage
of  grapefruit and oranges grown.  A great deal of the fruit is
grown  in  the  "ridge" area, a region of  sand  hills  running
north-south  in the center of the state.  Much is also grown in
the  "flatwoods" area, a region of coastal lowlands which flank
the  ridge area on the east and west coasts.  This entire  area
was  scrutinized  in terms of four primary groups  of  factors,
differences  in  which,  it was felt, would  cause  substantial
differences  in  the  fate  and transport of  aldicarb  in  the
unsaturated zone.  These groups were:

    1)   climate
    2)   soils
    3) -  management practices, and
    4)   thickness of the unsaturated zone.

Critical  variables  in terms of climate were preciptation  and
pan  evaporation depths.  These are factors which determine the

-------
     ALABAMA
                                                              Jacksonville
                                                                    Atlantic Ocean
                  3-st. Lucie
                  4-Indian River
                  5-Orange
                  6-Hardee
                  7-Martin..
                  B-Hillsborough
                  9-Highlands
                 10-De Soto
                 11-Paseo-
                 12-Hendry
                 13-Osceola
                 14-Brevard
                 15-Hanatee
                 16-Palm Beach
                 17-Harion
                 18-Volusia
                 19-Collier
                 20-Okeechobee
                 21-Lee
                 22-Hernando
                 23-Seminole
                 24-Charlotte
                 25-Glades
                 26-Pinellas
                 27-Putnam
                 28-Broward
                 29-Sumter
                 30-Sarasota
                 31-Citrus
                 32-Flagler
                 33-Machua
                 34-St. Johns
Figure   1.1  Florida  Counties Ranked by Citrus Acreage

-------
quantity of water percolating downward through the soils.

In the soil, the variables:

    o    field capacity water content (and saturated hydraulic
         conductivity),
    o    soil organic matter content, and
    o    soil pH

were  critically evaluated.  While field capacity water content
(and  hydraulic  conductivity) determine the velocity at  which
water  moves  through  the soil, soil  organic  matter  content
affects  chemical  adsorption  and  soil  pH  strongly  affects
aldicarb   degradation.     It  was  determined,  through   data
analysis,  that  differences in these factors  were  pronounced
among  four  soil orders on which citrus is grown  in  Florida;
the entisols, ultisols,  spodosols and alfisols.

Management  practices  were  also  investigated.   The  primary
difference  in  management which affects aldicarb migration  is
the  type  of  irrigation water  application  system  utilized.
These are four dominant types of systems;

    o    overhead methods  (i.e., permanent set or traveling
         volume guns)
    o    low volume spray
    o    flood or seepage, and
    o    drip or trickle.

In  addition,  the practice of bedding citrus in the  flatwoods
areas  is a dominant  edaphic practice.  This is done to  create
a  greater  root  zone  depth  for the  trees  and  to  promote
drainage of the root  zone.

In  the  flatwoods areas, water tables are typically very  close

-------
to  the  surface and drainage of the root zone is essential  to
production.    In   the  ridge  areas,  on  the   other   hand,
unsaturated  zone  thicknesses are greater and  such  practices
are unnecessary.

Based   on   climatic,  soils,  management  and   physiographic
considerations,   the   Florida  citrus  growing   region   was
subdivided  into the six areas shown in Figure 1.2.  Area 1  is
an  area of primarily ultisols while area 2 consists  primarily
of  entisols.  In these areas, no-irrigation, overhead  methods
and  low volume spray and drip methods are utilized and  citrus
is  not  bedded.  Areas 3 and 5 consist primarily of  spodosols
while  areas  4  and 6 contain primarily  alfisols.   In  these
areas,  citrus  is bedded and the overhead, flood,  low  volume
spray  and drip methods are used for irrigation.  Areas 5 and 6
receive  about 10% more rainfall on an annual basis than  areas
3 and 4, however.

At  the outset of the modeling exercise several decisions  were
made   concerning   the  possible  scenarios  which  could   be
simulated.    First,   the  drip  irrigation  method  was   not
simulated for the following reasons:

    1)   From the geometry of the wetted area and the treated
         aldicarb  band it was determined that, in most  cases,
         the  two  would not intersect.  There is  a  potential
         for overlap in very young groves, however.

    2)   Because PRZM operates on a daily timestep, and, due to
         the  fact  that  two to three water  applications  are
         needed  per  day  during peak periods,  this  type  of
         system could not be adequately simulated.

Second,  because  PRZM is a one-dimensional model, it was  felt
that  meaningful  simulation  of flood irrigtion would  not  be
                              8

-------
                                                       Jacksonville
                                                           Atlantic Ocean
Figure  1.2  Subareas delineated for unsaturated  zone modeling,

-------
possible, therefore, this irrigation method was also dropped.

Therefore,  a  total  of twenty-four different  scenarios  were
actually   simulated  for  the  unsaturated  zone.   These  are
indicated  in the matrix of Figure 1.3.  Ultimately, hydrologic
simulations  showed  that the attempted representation  of  the
low  volume spray method was also incorrect.  Therefore,  these
simulations  were  not used in the final analysis of  pesticide
loadings to the saturated zone.

Simulation  of  aldicarb  transport under  "freeze  protection"
water  application events using PRZM was also ruled out because
of  the daily simulation timestep.  Accurate representation  of
this  practice would require at least hourly inputs (e.g.,  air
temperature) and an hourly simulation timestep.

1.3.1.2  Saturated Zone—

The  occurrence  and  movement of ground water  in  the  citrus
growing  area  of Florida are closely related to  its  geology.
The   central  Floridan  peninsula  is  comprised  of  a  thick
sequence   of  hydrologically  connected  limestone  formations
which  make  up  the principal artesian aquifer,  the  Floridan
Aquifer.   This  is  overlain by younger  alluvial  and  marine
deposits  which  contain the unconfined surficial  aquifer  and
the  intermediate confined aquifers.  All three are present  in
the  citrus  growing areas and supply drinking water.  For  the
purposes  of modeling, this very complicated system of aquifers
was simplified.

Figure   1.4   shows  the  delineation of  the  typical  aquifer
geometries  considered  in this study.  Areas 1,  2, and   3  are
areas  that  include  the Floridan Aquifer system.  Area  1  is
where  the Floridan is unconfined and considered  alone.  Area  2
is  where a surficial unconfined aquifer overlies the  Floridan
                             10

-------
                           HIGH
                      NONE
                           OVERHEAD
   LOW
VOLUME SPRAY
                                                             LOW
                                                 NONE
                                                      OVERHEAD
   LOW
VOLUME SPRAY
              THICK
 ENTISOLS
 ULTISOL8
 SPODISOLS
              THIN
              THICK
              THIN
              THICK
              THIN
              THICK
 ALFISOLS
              THIN
Figure 1.3   Matrix showing  unsaturated  zone modeling scenarios
              simulated.

-------
        ALABAMA
                                          Jacksonville
                                                Atlantic Ocean
  1 -


  2 -



  3 -




  4 -


  5 -
      Explanation

Floridan Aquifer Alone
  -  Unconfined
Leaky Two-Aquifer System •
  -  Surficial Aquifer
.  -  Floridan Aquifer

Leaky Three-Aquifer System
  -  Surficial Aquifer
  -  Intermediate Aquifer
  -  .Floridan Aquifer
Surficial Aquifer Alone
  -  Unconfined
Two-Aquifer System
  -  Surficial Aquifer
  -  Intermediate Aquifer
Figure 1.4  Principal aquifer geometries in the
            citrus growing area.
                          12

-------
with  a  confining layer in between.  Together they comprise  a
leaky  two-aquifer  system.  In area 3, a three-aquifer  system
exists,  consisting  of the unconfined surficial aquifer and  a
confined intermediate aquifer overlying the Floridan Aquifer.

In  areas  4 and 5 the Floridan is not included because of  its
poor  water  quality or because its potentiometric  surface  is
above  the  water table.  Area 4 is where only  the  unconfined
surficial  aquifer  is  considered.  Area 5  includes  both  an
unconfined   surficial  aquifer  and  a  confined x intermediate
aquifer.

Table  1.1  shows the approximate citrus acreage in each  area.
The  largest  amount  of citrus is grown in Area 4,  where  the
surficial  aquifer  is considered alone.  This is  followed  by
Areas  2,  3,  and 5 where a multi-aquifer system  is  modeled.
Area  1  has the smallest amount of citrus grown   (only  3.0%),
yet   it   is  important  to  consider  because  the  risk   of
contamination  is  high.   Therefore  none of  the  areas  were
eliminated for lack of significant quantity of citrus.

These  five  geometries are grouped into three general  aquifer
configurations  for  modeling:   1) the Floridan, as  a  single
unconfined  aquifer  (Area 1), 2) the unconfined surficial (Area
4),  and  3) a general two-aquifer system (Areas 2, 3, and  5),
be  it  the  surficial  aquifer overlying the  Floridan  or  an
intermediate   confined   aquifer.    Each   of   these   three
configurations  were simulated with a worst and an average  set
of hydrogeologic parameters.

In  addition to the physical and chemical properties of aquifer
materials  and  their effects on chemical fate  and  transport,
the  drinking water well influences the movement of the  ground
water  and,  therefore,  the transport of aldicarb.   The  well
position  relative  to the pesticide source area,  the depth  of

                             13

-------
TABLE 1.1  APPROXIMATE CITRUS ACREAGE ASSOCIATED WITH EACH
           AQUIFER GEOMETRY
  County
( in ranked
order)
Polk
Lake
St. Lucie
Indian River
Orange
Hardee
Martin
Hillsborough
DeSoto
Pa sco
Hendry
Osceola
Brevard
Manatee
Palm Beach
Marion
Volusia
Okeechobee
Lee
Hernando
Seminole
Charlotte
Glades
Pinellas
Sumter
Citrus
Total
%
Area
123
31,000 63,000
87,000

10,000
28,500
42,000

14,000 17,500
16,500
13,000 17,000

12,000

8,000

2,000 2,000
2,500
2,000

5,000 700
600


2,000
600 1,000
1,000
21,600 212,300 147,000
3.0 29.5 20.4

4

17,000
64,000
50,000
14,000

32,000
3,500
13,000


2,500
15,000
5,000
7,500

4,500

5,000

5,000
4,500


1,300

243,800
33.8

5
31,000

7,000



6,500

3,500

30,000
2,500


5,000


4,500
1,000


1,000
4,000



96,000
13.3
                              14

-------
the  well  and  the rate at which it pumps  all  influence  the
ground-water flow regime.

The  distance of a drinking water supply well from the area  of
pesticide   application  is  crucial  in  the  evaluation   and
assessment  of  potential  human exposure to  aldicarb  through
drinking  water.  The greater the distance necessary to  travel
from  the source area to the well through the ground water, the
smaller  the chance of contamination.  The distance of the well
from  the source is an important management consideration.  The
current  EPA standard prohibits the use of aldicarb within 91 m
(300  ft) of a water supply well.  This distance was  evaluated
in  all the scenarios.  The EPA is considering 300 m (1000  ft)
as  an  alternative  distance for  aldicarb  regulation.   This
distance was also simulated in all the scenarios.

In   the   Florida  citrus  growing  area,  water  wells   vary
considerably   in   depth  depending  on  the   local   aquifer
characteristics   and  water  needs.   Deep  wells  with  large
capacities   are  usually  drilled  to  supply  municipalities.
Shallow  wells  have  much smaller capacities  and  supply  the
small  domestic user.  The depth that water is withdrawn from a
well  influences the localized aquifer flow system.  A  shallow
well  affects  the  near surface flow  system  and,  therefore,
would  influence the movement of a leaching contaminant  sooner
than a deep well with an equivalent pumping rate.

The  well rate determines the magnitude of the influence on the
regional   flow  system.   A  low  pumping  rate  has  a  small
localized  effect.   High  pumping rates can  change  the  flow
system  on  a much larger scale, greatly increasing the  ground
water  velocities  towards  the  well.   Two  well  rates  were
simulated with every well in each hydrologic setting.

The  physical  factors   influencing aldicarb transport  in  the

                              15

-------
saturated  zone  are reviewed in Table 1.2 with the cases  that
were  considered  for  each  factor.   To  recapitulate,  three
aquifer  geometries  were considered, the  unconfined  Floridan
Aquifer,   the  unconfined  surficial  aquifer  and  the  leaky
two-aquifer  system.   Each  of  these  three  geometries  were
simulated  with  a  worst and an average set  of  hydrogeologic
parameters.   A  drinking water well was modeled at both  91  m
(300  ft)  and 300 m (1000 ft) distances.  Two depths  and  two
pumping  rates  were simulated in each  hydrogeologic  setting.
All  the combinations of these factors resulted in  forty-eight
modeling  scenarios.   This  number  doubled  to  96  potential
scenarios  because  of the consideration of two chemical  decay
rates.   Two  decay rates were used; 1) the best estimate of  a
representative rate, and 2) no decay.

The  scenarios  actually simulated are shown in the  matrix  in
Figure   1.5.   Forty-four  simulation  runs  were  made.   The
motivation  for  dropping most of the scenarios which were  not
run  was  lack of significant (> 1 ppb) contamination in  cases
where worse concentrations would have been expected.
1.3.2  Pesticide Characteristics
Aldicarb   is   a   nematicide,  acaricide,  and   a   systemic
insecticide.   Its  environmental  fate  is  dominated  by  two
factors:   the fact that it forms two toxic daughter  products,
and  its  high  mobility in soils.  Degradation  of  the  toxic
residues  of the compound is of intermediate duration  compared
to other pesticides.

Aldicarb  is  a white crystalline solid which  is  incorporated
into  soil  as  a granule containing either 10% or  15%  active
ingredient.   In  order  to be effective, it must  dissolve  in

                              16

-------
   TABLE  1.2  CASES CONSIDERED FOR SATURATED ZONE MODELING
Influencing Factors
Cases Considered for Each Factor
Aquifer geometry
1) Unconfined Floridan Aquifer
2) Unconfined surficial aquifer
3) Leaky two-aquifer system
Aquifer properties
1) Worst case
2) Intermediate case
Well distance
1) 300 ft
2) 1000 ft case
Well depth
1) Shallow
2) Deep
Well rate
1) High rate
2) Low rate
                              17

-------
00
             SURFICIAL
             AQUIFER
            UNCONFINED
            UNCONFINED
             FLORIDIAN
             AQUIFER
            TWO-AQUIFER
              SYSTEM
                           AVERAGE
                           WORST
                           AVERAGE
                           WORST
                           AVERAGE
                           WORST
                                        DECAY
NO. DECAY
                                        DECAY
                                       NO DECAY
                                        DECAY
NO DECAY
 DECAY
                                       NO DECAY
                                        DECAY
NO DECAY
                                        DECAY
                                      NO DECAY
                                                            HIGH
                                                                  LOW
                                                                         HIGH
                                                                               LOW
                                                                                     HIGH
                                                                                           LOW
              Figure  1.5   Matrix  showing  saturated zone scenarios  simulated.

-------
water.   Once  this  happens in soils, the compound  begins  to
transform and degrade.

The  current theory is that aldicarb is fairly rapidly oxidized
to  aldicarb sulfoxide which in turn is more slowly oxidized to
aldicarb  sulfone.   Concurrently, these three  carbamates  are
transformed, by hydrolysis to corresponding oximes.  Hydrolysis
is  a  chemical  reaction in which water breaks up  an  organic
molecule  (RX),  such as aldicarb, by breaking a carbon-X  bond
and replacing it with OH from the water molecule.

These  products of hydrolysis are far less toxic than aldicarb,
its  sulfoxide  or its sulfone and are of little  environmental
concern.   A  schematic of these processes is shown  in  Figure
1.6.

The  investigation  of  the  rates  and  coefficients  used  to
describe  fate  and  transport  was also  divided  between  the
unsaturated and saturated zones as described below.

1.3.2.1  Unsaturated Zone—

PRZM  was modified in this study so that the transformation  of
aldicarb  to  its daughter products could be simulated  in  the
unsaturated  zone.   Therefore, the rates k^ through k5 had  to
be  estimated  for each transformation/degradation pathway  and
adsorption  partition  coefficients  had to  be  estimated  for
aldicarb,   its   sulfoxide   and  its   sulfone.    This   was
accomplished   by   taking  rates  from  the   literature   and
regressing  the  values  on levels of  environmental  variables
such  as pH, temperature, soil water content and organic  water
at  which each study was conducted.  Then, by using local  soil
environmental  conditions  for  each  soil  order,  rates  were
predicted    for    each   scenario.    Adsorption    partition
coefficients  were determined using average values of Koc found

                             19

-------
                  Aldicarb
      Aldicarb Sulfoxide
Aldicarb Sulfone
       CH3

CH,S - C - CH
  3    i
                            O
                            II
                          NOCNHCH.
       O  CH,
       II   I 3
   •CH-jS - C  - CH

          CH.,
  0      *2   O   CH.
  II           II    l 3
NOCNHCH3—- CH-jS - C  - CH

              O   CH-
             NOCNHCH.
                              (Hydrolysis)
                        (Hydrolysis)
to
O
Nontoxic Oximes and  Nitrites
               Figure  1.6 Schematic  of  aldicarb environmental chemical Pathways

-------
in  the  literature  with  localized  values  of  soil  organic
carbon.

The  values  of the rates and coefficients were then used in  a
modeling  verification study.  Four simulations were  performed
at  three different sites in Florida for which residue sampling
data  was available.  This data was collected in 1983 and 1984.
Simulations  indicated  that  rates and coefficients  used  for
spososols  (i.e.,  flatwoods soils) were well estimated by  the
multiple  regression  equations  derived  from  the  literature
data.   Simulations  of entisols (i.e., ridge soils),  however,
indicated  that  degradation  rates  used in  this  study  were
probably too fast.

1.3.2.2  Saturated Zone—

For  saturated  zone  simulations  the CFEST  model  was  used.
CFEST  simulates the transport of a single constituent which is
subject  to  adsorption and degradation.  Therefore,  a  single
first  order  decay  rate  had to  be  derived  for  simulation
purposes.   Since  the principle degradation pathway in  ground
water  is  thought  to  be  hydrolysis,  literature  values  of
hydrolysis   rates  were  analyzed.   Using  values  of   these
hydrolysis  rates  at  various levels of  pH  and  temperature,
activation  energies for the hydrolysis reaction for  aldicarb,
aldicarb  sulfoxide and aldicarb sulfone were calculated  using
the   Arrhenius  equation.   Then  using  localized  values  of
aquifer  pH and temperature in Florida, first-order degradation
rates  were  determined  for specific scenarios.   Because  the
predicted  rates for sulfone were fastest, these were used  for
scenarios   in   which  decay  was  simulated.    As   depicted
previously  in  Figure  1.5,  scenarios  were  also  run  using
no-decay    of    the   contaminant.    Adsorption    partition
coefficients   were  set  to  zero  in  the   saturated   zone
                              21

-------
simulations  as  no organic matter is thought to be present  in
aquifer materials.

In  conducting  the literature survey on degradation  rates  it
was  discovered  that other porous media properties may  affect
aldicarb   degradation.   Recently  much  higher  disappearance
rates  for  aldicarb sulfoxide and aldicarb sulfone  have  been
reported   when  incubated  under  anaerobic   versus   aerobic
conditions.    Under  anaerobic  conditions,  disappearance  of
these  two compounds may be from 8 to 100 times faster than  in
the  same soil under aerobic conditions.  The current theory is
                            2+
that   the  presence  of  Fe    in  solution  catalyzes   these
reactions.

Very  little  specific data is available on redox potential  or
                             2+
dissolved  ferrous  iron  (Fe  ) in Florida.  The  most  common
                                                          2+
form  of  iron  in  the ground water is ferrous  iron  (Fe   ),
however,  and ground water with a pH between 6 and 8 can  carry
as  much  as  50  mg/1 of ferrous  iron  at  equilibrium.   The
occurrence  of  1.0  to  10 mg/1 of iron  in  ground  water  is
common.   Dissolved iron in the ground water in Florida  varies
from  as little as 0.01 mg/1 to as high as 20 mg/1.  In general
the  surficial  aquifer has higher iron concentrations  because
the  aquifer materials contain more iron-bearing minerals  than
the  predominantly  limestone  Floridan Aquifer  (Irwin,  1984,
personal communication).

If  the  transformation  of  aldicarb is  indeed  catalyzed  by
ferrous  iron,  the  process would be important in  the  ground
water  in  Florida.   At this time, the process  is  so  poorly
understood  it was not included in the quantitative analysis of
the pesticide parameters for modeling.
                              22

-------
1.4  SUMMARY OF RESULTS
This   section   overviews   the  important  results   of   the
unsaturated  and saturated zone modeling and calculation of the
resulting    well   water   concentrations.    More    detailed
information is contained in Section 4 of this report.

Important  to this discussion is the fact that the  unsaturated
and  saturated  zone modeling efforts were essentially  pursued
independently.   Fourteen years of meteorologic data were  used
to  drive the unsaturated zone model.  The resulting output was
a  time  series  of  pesticide  loadings  at  the  top  of  the
saturated  zone  (i.e.,  ground water table).   Therefore,  the
unsaturated  zone simulations were fully dynamic, both for flow
and  transport.   For PRZM, which runs very economically,  this
type   of  simulation  was  feasible  for  a  large  number  of
scenarios.

With   CFEST,   steady-flow,  unsteady  contaminant   transport
simulations   were   used,  for  several  reasons.   First,   a
two-dimensional  and  sometimes three-dimensional geometry  was
needed  which  required  far greater time and  money  resources
than   for  the  one-dimensional  PRZM.   Long,  unsteady  flow
simulations   were  impractical,  given  the  large  number  of
scenarios  to  be run.  Second, sensitivity  simulations  using
CFEST  showed  that transient flow conditions caused by  widely
fluctuating  recharge  rates  from the  unsaturated  zone  were
minor   and  short-lived.   This  conclusion  allowed  for  the
convenient   use  of  steady  flow  defined  by  the   regional
hydraulic  gradient  and the assumption of a fixed  depth  from
the land surface to the saturated zone.
                              23

-------
1.4.1  Unsaturated Zone
The  resulting annual pesticide loadings from the sixteen final
unsaturated  zone  scenarios  were analyzed  in  several  ways.
First,  geometric  mean pesticide loadings from  all  scenarios
were  subjected  to  one way analysis of  variance  (ANOVA)  to
determine  if  any substantive differences occured between  the
mean  loadings of the various scenarios.  These analysis showed
that  of the means of the sixteen scenarios simulated the  only
significant  differences  were among thick ridge  soils  (i.e.,
entisols  and ultisols), thin ridge soils, and flatwoods soils.
Visual  inspection  of the frequency distribution  of  annually
leached  pesticide  loads confirmed this.  Thus, the output  of
the  sixteen scenarios run were condensed into three scenarios;
"ridge"  areas with thick unsaturated zones, "ridge" areas with
thin unsaturated zones, and "flatwoods" areas.

By  condensing the information in the sixteen scenarios down to
three,  better probability estimates are obtained.  Figures 1.7
through  1.9  show the three resulting  frequency  distributions
for  the  thick  unsaturated zone ultisols and  entisols,  thin
unsaturated  zone  ultisols and entisols, and the  alfisol  and
spodosol groupings.

These  loads, however, are leached only from the treated  band.
Much  of  the  area in the typical citrus grove  is  untreated,
with  no  resulting  leached load.  Therefore, the  loads  from
Figures  1.7  through  1.9  must be reduced  by  the  ratio  of
treated   to   total  area  in  the  grove.   This   ratio   is
approximately  0.32  to  1 for ridge citrus and 0.20 to  1  for
double bedded flatwood citrus.

Table  1.3 summarizes the pesticide loads for the 90, 50 and 10
percentile   exceedance  probabilities  for  the  three   final

                             24

-------
                            RIDGE SOILS


                         THICK UNSATURATED ZONE
0
UJ
Q
UJ
ui
u
X
UJ

m

o

o

o
UJ


o
o
m
<
to
o
cc
a.
   0.00

     1.E
-5      1.E-4       1.E-3      1.E -2


                 PESTICIDE LOAD IN KG/HA
1.E-1
1.E 0
  Figure 1.7
        Frequency of annual quantity of pesticide

        leached to the  saturated  zone from thick

        Entisols and Ultisols.
                              25

-------
UJ
Q
UJ
Ul
O
X
tu

CO
Q
UJ
o
5
z

H

X
m
<
o
O
cr
Q.
                          RIDGE SOILS

                       THIN UNSATURATED ZONE
   0.00
     1.E
-5      1.E-4       1.E-3      1.E -2


                 PESTICIDE LOAD IN KG/HA
1.E-1
1.E 0
  Figure  1.8 Frequency of  annual quantity of pesticide
              leached to the  unsaturated zone from thin
              Entisols and  Ultisols.
                              26

-------
                        FLATWOODS SOILS

                      SPODOSOLS AND ALFISOLS
    1.E-5      1.E-4      1.E-3       1.E -2

                        PESTICIDE LOAD IN KG/HA
                                 1.E-1
1.E 0
Figure  1.9
Frequency of annual  quantity of pesticide
leached to the saturated zone from Spodosols
and Alfisols.
                               27

-------
 TABLE 1.3 SUMMARY OF PESTICIDE LOADINGS PER UNIT CITRUS BLOCK
           AREA FOR THE THREE FINAL UNSATURATED ZONE SCENARIOS
                                   Exceedance Probability
Scenario                       0.90         0.5U         0.10
Alfisols and spodosols        1.8E-5       4.0E-4       2.2E-3


Thick ultisols and entisols   6.4E-6       4.2E-4       3.2E-3


Thin ultisols and entisols    1.9E-3       1.3E-2       9.6E-2
                             28

-------
unsaturated  zone scenarios.  These loads have been  multiplied
by  the  above  ratios  to yield loads per  unit  citrus  block
instead  of unit treated area.  A unit citrus block is  defined
as  the  area alloted to each tree according to the spacing  in
the  grove.   The  table reveals that lowest  loadings  are  in
general  associated  with  alfisols and  spodosols.   For  this
scenario  there  is only a 10% probability that  the  pesticide
load  leached  to ground water will exceed 0.002 kg/ha.   Loads
from  thick entisols and ultisols exceed those for the alfisols
and  spodosols  slightly.   The highest loadings  emanate  from
thin  unsaturated  zone entisols and ultisols.  There is a  10%
chance  that loads to ground water will exceed 0.1 Kg/ha.   The
thickness  of  the unsaturated zone in this scenario is 180  cm
for  ultisols and 270 cm for entisols and the input load is 5.6
Kg/ha  or  5 Ib/acre.  Increasing or dereasing the load in  any
of  these scenarios by a ratio 'x', would result in an increase
or  decrease  in the load by the same ratio.  For instance,  if
the  application  rate  were  doubled from 5.6  Kg/ha  to  11.2
Kg/ha,  the simulated load at the 10% exceedance level in  thin
entisols  and  ultisols would also double, from 0.096 Kg/ha  to
0.19 Kg/ha.

Also  of  interest is the quantity of aldicarb and that of  its
two  toxic  metabolities  in  the  leached  load.   Simulations
revealed  that  usually less than one percent of  the  aldicarb
parent  is  leached to the saturated zone under  any  scenario.
Under  the  thin unsaturated zone ultisols and entisols,  about
60%  aldicarb  sulfoxide and 40% aldicarb sulfone makes up  the
leached  load  (a  1.5 to 1 ratio).  In the  thick  unsaturated
zone  entisols and ultisols, the ratio is closer to 0.17 to  1,
sulfoxide  to  sulfone.  The spodosols and alfisols, the  ratio
is roughly the same, 0.19 to 1.

Obviously,   since   these  transformations   are   kinetically
controlled  the quantity of aldicarb and its toxic metabilities
                             29

-------
appearing  in the leachate is a function of the residence  time
of  the chemical in the profile.  The sooner after  application
the  pesticide  is  leached  to the  saturated  one,  the  more
aldicarb and aldicarb sulfoxide will appear in the leachate.
1.4.2  Saturated Zone
A  summary  of the results of the saturated zone scenarios  are
shown  in  Table  1.4.  This table shows the 90,  50,  and  10%
probabilities  of exceeding the given relative  concentrations.
These    "relative"    concentrations   are    the    simulated
concentrations   at   the   well   divided   by   the   initial
concentration  in  the  recharge water and  are  dimensionless.
Because  a  unit load (1 kg/ha) was used for the input  in  all
model  simulation  runs, all the well-water concentrations  are
expressed in this manner.

In  general  the  surficial  aquifer  cases  show  the  highest
relative  concentrations  followed by the worst cases  for  the
Floridan  Aquifer  and then the average cases for the  Floridan
and  the  two-aquifer  cases.  The surficial aquifer  has  high
relative  concentrations for two reasons.  The relatively  thin
aquifer  has  less  water for the pesticide to  disperse  into.
Also,  the  well-induced gradients and permeabilities are  high
enough  to significantly increase the ground-water velocity  to
the   well.   This  allows  less  time  for  decay.   The  high
concentrations  determined for the worst case scenarios in  the
Floridan  Aquifer  are  primarly due to the  high  ground-water
velocities  which  leave little time for decay  or  dispersion.
The  two-aquifer  scenarios and the average Floridan  scenarios
have  slower  ground-water velocities and thick  aquifers  that
dilute the initial pesticide concentrations.
                              30

-------
TABLE  1.4   RELATIVE CONCENTRATION VALUES FOR
              INDICATED SATURATED  SCENARIOS AT  THREE
              EXCEEDANCE PROBABILITIES
Scenario



Floridan Aquifer - worst  Case

  91 meters to well
    Deep well, no decay
    Decay
    Shallow well, no decay
    Decay

  300 meters to well
    Deep well, no decay
    Decay
    Shallow well, no decay
    Decay
    Exceedence Probability
 .90         .50         .10
6.4E-4
8.1E-5
4.0E-3
1.3E-4
1.1E-3
3.3E-5
4.0E-3
3.3E-5
4.4E-3
3.7E-4
l.OE-2
5.8E-4
5.0E-3
1.4E-4
8.5E-3
1.8E-4
8.6E-3
1.7E-3
1.5E-2
2.6E-3
8.4E-3
6.2E-4
1.2E-2
6.4E-4
Floridan Aquifer -  average case

  91 meters to well
    Deep well, no decay              5.7E-4      3.3E-3      7.1E-3
    Decay                           2.5E-6      1.8E-5      4.5E-3
    Shallow well, no  decay           6.3E-4      3.6E-3      7.5E-3
    Decay                           2.5E-6      1.8E-5      4.4E-5

  300 meters to well
    Deep well, no decay              1.6E-4      7.3E-4      3.0E-3
    Deep well, decay                 9.2E-9      3.1E-8      4.6E-8
    Shallow well, no  decay           9.6E-5      4.1E-4      1.9E-3
    Decay                           2.6E-9      8.7E-9      1.3E-8


Surficial Aquifer - worst case
  91 meters to well
    High pumping rate, no decay      1.1E-2      5.0E-2      7.3E-2
    Low pumping rate, no decay       2.8E-2      1.3E-1      1.7E-1
    Both, decay                     2.1E-4      1.1E-3      3.3E-3

  300 meters to well
    High pumping rate, no decay      5.4E-3      2.6E-2      5.9E-2
    Low pumping rate, no decay       6.7E-3      4.7E-2      l.OE-1
    Both, decay                     4.0E-6      1.8E-5      3.3E-5


Two-aquifer System, worst case
  91 meters to well
    High pumping, no  decay           1.5E-4      6.9E-4      1.7E-3
    Low pumping, no decay            3.4E-4      1.1E-3      2.1E-3
    Both, decay                     1.3E-5      6.5E-5      2.7E-4
                                31

-------
Other  trends  can be observed from the results.  As  expected,
the   cases   modeled   with   decay   resulted   in   relative
concentrations  much  lower than those modeled with  no  decay.
The  greatest  difference  is in cases where  the  ground-water
velocity  is  very slow.  On the other hand, in cases like  the
Floridan  worst  case the relative concentrations are not  much
lower when decay is simulated.

The  pumping  rate of the well has little influence.   This  is
especially  true  in  cases  where  the  regional  gradient  is
dominant  and where decay is simulated but even in cases  where
the  well-induced gradient is dominant like the surficial worst
cases  the  relative concentrations do not differ by much.   In
the  Floridan  worst  cases  the  greatest  difference  due  to
pumping  rates is 3%.  The surficial worst cases with no  decay
show  differences  as  much  as 50% between the  high  and  low
pumping  rates,  but  even in this case, the  trends  are  very
similar.

Concentrations  are  higher  in wells only 91 m away  from  the
source  area.   In  the  cases  with  no  decay,  the  relative
concentrations  at the 300 m well are smaller because there  is
more   dispersion.   The  Floridan  worst  cases  show  only  a
difference  of about 20% between the 91 m and the 300 m  cases.
For  the  average Floridan and surficial worst cases,  relative
concentrations  differ  by as much as 70%.  In the  cases  with
decay,  the  relative  concentrations are usually at  least  an
order  of magnitude smaller at the 300 m well than at the 91  m
well.   Again, the greater amount of time required to travel to
the 300 m well, allows more time for decay to occur.
                              32

-------
1.5  COMBINED RESULTS OF UNSATURATED AND SATURATED ZONE
     MODELING
1.5.1  Physical Overlap of Saturated and Unsaturated Zones
The   results  from  the  unsaturated  zone  modeling  and  the
saturated  zone  modeling  efforts were  combined  to  evaluate
expected  aldicarb concentrations in drinking water wells.  The
unsaturated  zone scenarios cover two physiographic areas:   1)
areas  where  entisols and ultisols occur, and 2)  areas  where
spodosols  and  alfisols occur.  Even though for  the  entisols
and   ultisols  both  thin  and  thick  unsaturated  zones  are
considered,  thick and thin areas occur randomly and  therefore
no  attempt is made to separate them physically.  The saturated
zone  is comprised of three physical aquifer systems that  were
modeled:    1)   the  unconfined  Floridan  Aquifer,   2)   the
unconfined  surficial aquifer and 3) the multi-aquifer  system.
Figure  1.10 shows the areas of overlap of the unsaturated  and
saturated  zone  scenarios.  The entisols and  ultisols   (ridge
soils)  overlap all three of the saturated aquifer system.  The
spodosols  and  alfisols  (flatwoods soils)  overlap  with  the
surficial  aquifer  and the multi-aquifer systems but not  with
the  unconfined  Floridan  Aquifer.   The results  of  the  two
modeling  studies  were  combined in all  cases  where  overlap
occured.
1.5.2  Results
The  complete set of results showing well-water  concentrations
at  the  10  and  50  percent exceedance  values  for  all  the
scenarios   are  shown  in  Appendix  E.   These  results   are

                             33

-------
                                            Jacksonville
                                                Atlantic Ocean
                                          FLATWOODS
                                            SAILS.,
         LEGEND



FLORIDIAN AQUIFER - UNCONFINED


MULTI - AQUIFER SYSTEM


SURFICIAL AQUIFER - UNCONFINED
Figure 1.10   Overlap  of unsaturated  and saturated
               model  areas.
                          34

-------
summarized  in  Table 1.5 which reports the highest  calculated
concentration  for  each catagory shown.  In general the  well-
water  concentrations are very low.  The highest  concentration
is  6.5  parts per billion (ppb) in a scenario  without  decay.
This  is  just at the detection limit of 5 to 6 ppb  for  total
toxic   residues  (Rao,  1984,  personal  communication).   The
highest  concentrations  for  scenarios with decay are  in  the
order of 10   ppb.

The   well-water  concentrations  show  three  general  trends.
First,  the  highest  concentrations are associated  with  thin
ultisols  and  entisols,  followed by the  thick  ultisols  and
entisols  and  the  alfisols  and spodosols.   Second,  in  the
saturated  zone,  the  worst cases for  the  surficial  aquifer
generally  have  the highest well water concentrations  closely
followed  by  the unconfined Floridan worst cases  and  finally
the  worst cases of the two-aquifer system and average Floridan
cases.   Third,  within each set of saturated  zone  scenarios,
the  highest concentrations result from the no decay simulation
with  the well 91 m  (300 ft) from the source area, followed  by
no  decay  simulated with a well at 300 m  (1000 ft).  Next  are
the  decay  scenarios  with  the well at 91  m.   The  smallest
concentrations  result  from  the decay simulations  where  the
well is 300 m from the source.

1.5.2.1  Surficial Aquifer—

Combined  with  the  thin entisols and ultisols,  the  surfical
aquifer    worst   cases   generally   produce   the    highest
concentrations.   The highest value is for worst case hydraulic
parameters,  no  pesticide decay, with a shallow well at  91  m
distance  from  the  source area, pumping at a low rate,  having
a  well  water  concentration of 6.5 parts per  billion  (ppb).
With  decay  the  concentrations drop one  to  three  orders  of
magnitude.   When  the  well  is 300 m  from  the  source,  the

                              35

-------
TABLE 1.5  HIGHEST CALCULATED ALDICARB  CONCENTRATIONS  (in ppb)
           .IN THE GIVEN COMBINED UNSATURATED/SATURATED CATAGORIES

Floridan
Worst Cases
Floridan
Average
Surf icial
Worst Cases
Two-
Aquifer
Worst
Cases
300
ft
1000
ft
300
ft
1000
ft
300
ft
1000
ft
300
ft
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
Flatwoods
Soils
No
Overlap
No
Overlap
l.SxlCf1
2.9xlO~s
9.0xlO~2
2.7xlO~5
2.7x!0"S
3.5xlO~*
Thick
Ridge Soils
2.3xlO~2
4xlO~8
1.8xlO~2
9.8xlO~"
2.2xlO~2
1.3xlO~"
3.8xlO~3
1.4x!0"7
2.2xlO~'
4.2xlO~s
1.3X10"1
4.2xlO~5
3.9xlO~3
5.1x10'"
Thin
Ridge Soils
6.9X10'1
1.2X10'1
5.5xlO~a
2.9xlO~2
S.SxlO'1
3.9xlO~3
2.6X10"1
4. xlO~6
6.5
1.3X10"1
3.9
1.3xlO~3
1.2X10"1
1.5xlO~2
                                36

-------
surficial   cases   without   decay  still   show   significant
contamination,  but when decay is simulated the  concentrations
are  in  the  order 10   ppb or less.   The  other  unsaturated
zone  scenarios  show the same trends as the thin  ridge  soils
when  combined with the surficial aquifer worst cases with even
lower  concentrations.   The highest concentrations are 2.17  x
10    and  1.49  x  10    for the thick  ridge  soils  and  the
flatwoods soils, respectively.

1.5.2.2  Floridan Aquifer Worst Cases—

Well-water  concentrations in the Floridan worst cases are  not
as  high  as  those in the surficial worst  cases  except  when
decay  is modeled with the drinking water well at 300 m.   When
evaluating  these  results  with no decay, it is  necessary  to
realize  that  these  concentrations could be much  higher  for
wider   source   areas  (see  Section  4.2.4).    The   highest
concentration  in the scenarios without decay is 6.9 x 10   ppb
with  the  well  at 91 m (300 ft) and 5.5 x 10   ppb  with  the
well  at  300 m  (1000 ft).  The scenarios simulated with  decay
are  generally   one  order  of magnitude  less  than  the  same
scenario without decay.

1.5.2.3  The Floridan Aquifer Average Cases—

The  well-water  concentrations in the Floridan  average  cases
vary  considerably  between  scenarios modeled with  decay  and
without  decay.   The concentrations calculated  for  scenarios
with  the  well  91 m  (300 ft) away from the source area and  no
decay  are  as high as 0.65 ppb.  In the cases with  decay  the
concentrations   drop over two orders of magnitude when the well
is  91  rn from the source and four to five orders of  magnitude
when the well is 300 m from source.
                              37

-------
1.5.2.4  Two-Aquifer System—

The  scenarios modeled with the two-aquifer system exhibit some
of  the  lowest concentrations.  The highest value is .12  ppb.
With  decay simulated the concentrations drop about an order of
magnitude.
1.6  CONCLUSIONS AND RECOMMENDATIONS
The  results presented in the foregoing section are based  upon
model  representations of generalized, regional unsaturated and
saturated   zone   modeling  scenarios.   Such  a   study,   of
necessity,  overlooks  specific "special situations" which  may
occur  within these regions.  To ameliorate the effects of such
over-sights,  the  general approach was to attempt to  look  at
"worst  case"  values of parameters which could be selected  to
describe   these  generalized  scenarios  and  to  adopt  other
"average  case"  values  when worst case  values  gave  results
indicative  of high contamination levels.  This was not done in
all  cases.   High  contamination  levels  generally  were  not
found,   however,   even   under   "worst   case"   conditions.
Inevitably,  higher  concentrations that those  predicted  here
will  be  found at sometime in someone's drinking  water  well.
The  primary intent of this concluding section is to  highlight
the  caveats of this simulation study yielding an indication of
the  types  of special cases under which higher than  predicted
concentrations  might  be found.  Secondarily,  recommendations
on how to avoid these high concentrations will be stated.
                             38

-------
1.6.1  Unsaturated Zone
In  terms of the unsaturated zone simulations, several  caveats
are  in  order.   First,  a limited  period  of  meteorological
record  (14  yrs)  was used for the simulations.   During  this
period,  1965  to  1978,  the mean annual  rainfall  was  below
normal.  If a longer period had been used, better estimates  of
leaching  probabilities  would  have  been  obtained  and  more
extreme  years  might  have been included, giving  better  tail
probabilities  at  the  high  end  of  the  leaching  frequency
distributions.   The  longer  period was not used  because  pan
evaporation  data was not available and the algorithms in  PRZM
used  to  estimate pan evaporation from mean daily  temperature
consistantly underestimated the magnitude of that time series.

Mean  values  of  state variables such as soil  field  capacity
water   content,   pH,  organic  carbon,  etc.  were  used   as
representative  of  the unsaturated zone soils.   Estimates  of
variability   were  available  for  these  parameters  and,  in
retrospect,  it  may have been appropriate to chose values  of,
say,   mean  plus  or  minus  one  standard  deviation  in  the
direction  of the worst case, as representative.  It should  be
noted  that most of the soil characterization samples available
were  not  taken  in  citrus groves but  in  areas  of  natural
vegetation.   Soils in citrus groves might have higher  organic
carbon  due to maintanence of the upper profile to promote root
growth  and  thus  have higher water holding  capacity.   These
conditions  would be conducive to stronger pesticide  retention
and  more rapid decay.  Therefore, the best estimate parameters
for  soils  having  natural vegetation might be  biased  toward
higher pesticide leaching then cultivated citrus soils.

The  pesticide  parameters,  derived from regression  on  those
environmental   state  variables,  were  chosen  based  on  the
                              39

-------
regression  estimate  rates and constants using mean values  of
the  state variables.  Thus, best estimate values of soils  and
pesticide  parameters  were  used as opposed  to  "worst  case"
values.   The  fact that the estimated pesticide parameters  do
not  represent  worst-case  values is demonstrated by  the  too
rapid  simulation of pesticide degradation in 1983 and 1984  at
the  Lake Hamilton and Davenport sites in the PRZM verification
runs.   Pesticide  leachate loadings from these  "ridge"  soils
would  probably  be greater than that represented in the  model
simulations  due  to  the  relative  longevity  of  the  actual
field-monitored  residues.   Unfortunately, the 1984  Davenport
data  was not available early enough in the study to affect the
selection of pesticide parameters for modeling scenarios.

Probably  the  only  crop parameter that is  crucial  in  these
simulations  is  the rooting depth.  Representative values  (60
cm  for spodosols and alfisols, 150 cm for ultisols and 240  cm
for entisols) were used for each soil order.

Vertical   hydraulic  transmission  rates  for  spodosols   and
alfisols  were based on only one data set for a  representative
soil  type of each soil order.  The confidence to be placed  in
this   estimate  is  low  and,  unfortunately,  the   pesticide
loadings  simulated for these soil types are heavily  dependent
upon   this  transmission  rate.   Therefore,  less  confidence
should  be placed in these loading estimates versus those  from
the  entisols  and ultisols based on  hydraulic  considerations
alone.   For  the  ridge  soils,  the  "unrestricted  drainage"
option  was  used in PRZM which does not require estimation  of
this parameter.

The  inability to use PRZM to simulate irrigation methods other
then  the  overhead  types is unfortunate.   Loading  estimates
were  not  produced  for  low-volume spray  or  flood  methods.
Overhead   irrigation  did  cause  an  increase  in   pesticide
                             40

-------
leaching  in  our  simulations although this increase  did  not
appear   to  be  statistically  significant.   It  is  certain,
however,that  irrigation  water,  even when  properly  applied,
increases  the  opportunity for leaching by natural  rainfalls.
For  those  methods (i.e., low-volume spray and trickle)  where
water  is  supplied to the tree over a relatively  small  area,
leaching  due to irrigation will occur if the water application
rate  exceeds  the  combined  capacity of  the  tree  roots  to
extract  and  the  soil  to store the  water.   Concerning  the
overall   effect,  however,  this  enhanced  leaching  may   be
mitigated  by  the  fact that it is occurring  over  a  smaller
area.

At  this point, the effects of flood irrigation on the leachate
load  of pesticide are uncertain.  It is evident that saturated
conditions  in  the profile increases downward water  movement.
However,  during the drainage cycle, pesticide-laden water  may
move  laterally,  decreasing the opportunity for  leaching,  at
least  from  the  field area.  In addition, capillary  rise  of
water  induced  by  seepage irrigation may  actually  cause  an
upward  movement  of chemical and subsequent removal  by  plant
roots.   Some two-dimensional modeling of these systems  should
be done at some point, if possible.

Simulation  of the application of water via irrigation  systems
for  freeze protection was not accomplished.  Extensive changes
required  PRZM  and each of appropriate input data to  properly
simulate   this  process  caused  it  to  be  eliminated   from
consideration.   Application  of large quantities of water  for
this   purpose  would  obviously  move  large  quantitities  of
aldicarb  through  the system.  It is interesting to note  that
for  the Floridan worst case saturated zone scenario with decay
and  a  shallow well 91 m  (300  ft) downgradient of the  source,
if   all   of   the   5.61  Kg/ha   application   of   aldicarb
catastrophically  entered  the  saturated zone,  the  resulting

                             41

-------
simulated  peak well water concentration would only be 8.2 ppb.
For   the   surficial  aquifer  worst  case,  under  the   same
conditions,  the simulated peak concentration would be 8.4 ppb.
For   300   m  (1000  ft)  downgradient  distance,   the   peak
concentrations  drop  to 2.0 and 0.08 ppb for the Floridan  and
surficial aquifers, respectively.

Based  on  results  of  the unsaturated  zone  simulations  and
subsequent  discussion,  there are a number of  recommendations
which  can  be made for practices which will tend  to  diminish
loadings from the unsaturated zone:

    1)   Avoid the application of aldicarb in ridge areas where
         there  is a thin unsaturated zone  (i.e., a high ground
         water table).

    2)   Avoid the application of irrigation water in the
         treated  band in quantities that would cause  movement
         of water and pesticide past the crop root zone.

    3)   Avoid the application of irrigation water for freeze
         protection  after  the  application of aldicarb  to  a
         grove.
1.6.2  Saturated Zone
In   the  saturated  zone  modeling,  emphasis  was  placed  on
modeling  the  "worst"  case scenarios, for the  three  general
aquifer    configurations.     Even    so,    the    well-water
concentrations  could  be underestimated for  several  reasons.
First,  assumptions  were  made for the  simulations   (such  as
distribution  of  pesticide  loads) that affect  the  predicted
concentrations.   Second and more important are the  situations

                              42

-------
that  were not simulated due to limitations of the model or the
limitations in the regional scope of the study.

Because  the unsaturated and saturated zone modeling were  done
concurrently,  the  output  of the PRZM runs was  not  directly
used  as input for CFEST.  Assumptions were made concerning the
timing   of  the  loads  from  the  unsaturated  zone  and  the
distribution  of  that  load based on preliminary  results.   A
standardized  leaching  event of four months duration was  used
over  which  the annual pesticide load was evenly  distributed.
For  most of the scenarios this was a worst case estimate since
four  months was the minimum time required to leach over 90% of
the  pesticide on the average.  Even in an anomolous year where
most   of   the  pesticide  was  leached  in  one  month,   the
concentrations   would  go  up  by  only  a  factor  of   four.
Likewise,  assuming uniform loading throughout the four  months
disregards   the   high  spikes.   But  this  too   would   not
significantly  change the pesticide concentrations as shown  by
the  transient sensitivity simulation where the actual loadings
from  a PRZM output resulted in concentrations that were higher
by only 10% in one year and lower in another year.

Sensitivity  simulations  demonstrated  the importance  of  the
size    of   the   pesticide   source   area   on    well-water
concentrations.   As  the  time required to  travel  under  the
source  area  increases,  the ground-water  will  receive  more
leached  pesticide  residues  from the  unsaturated  zone.   In
cases  with  low decay or no decay, this can result  in  higher
pesticide  concentrations  at the well.  For the  extreme  case
where  there is no decay and no dispersion, if the width of the
source  area  were to double, well-water  concentrations  would
also double.

In   the   scenario  simulations  the  source  area  was   only
considered  to  be  up-gradient of the well.   The  sensitivity

                             43

-------
simulations  demonstrated that concentrations would increase as
much  as four times in situations where the well dominates  the
regional  gradient.  Combined with the other factors  discussed
above,   concentrations  could  be  as  much  as  an  order  of
magnitude higher than those predicted by the model.

In  this  study, the worst set of the  hydrogeologic  parameter
values  represented some of the more extreme situations in  the
ground-water  environment.   Still the solution scheme used  in
CFEST  assumes a porous media where these hydrogeologic  values
are  averaged  over  the aquifer.  The more  likely  situation,
especially   in  the  limestone  aquifers,  is  that   solution
cavities   riddle  the  aquifer  creating  a  situation   where
fracture   flow  is  dominant.   Large  fractures  or  solution
cavities  may  have  velocities many times  larger  than  those
measured,  while  the  surrounding  rock  may  be  more  nearly
impermeable.

This  is a situation that can not be simulated by CFEST and  is
a  current  topic  of active research.  This type  of  fracture
flow  could  lead  to  concentrations much  higher  than  those
simulated.

Another  potentially  dangerious situation is that of a  direct
conduit  to the aquifer.  These can occur naturally in the form
of  a sink hole or can be man-made in the form of a leaky well.
When  the potentiometric surface of the lower aquifer is  lower
than  the water table either a sink hole or a leaky well allows
contaminated  water  to flow directly into the  lower  aquifer.
These  localized situations were not modeled.  A best  estimate
of  pesticide concentrations under these conditions would be to
use  the  results  from scenarios simulated  in  an  unconfined
aquifer.

In  all  the scenarios the hydraulic gradient was that  of  the

                              44

-------
combined  regional  gradient  plus  the  drinking  water  well.
Florida  has  areas of heavy ground-water withdrawals for  both
irrigation  and  municipal supply that depend on  well  fields.
The  effect  of these localized induced gradients due  to  many
wells   pumping  together  was  not  considered.   Much  higher
gradients  would  make ground water velocities much higher  and
therefore allow less time for decay.

In  summary,  all  situations were not simulated due  to  model
limitations,  limitations  in the regional scope of the  study,
or   assumptions  made  for  the  purposes  of  the   modeling.
Considering  the  effects of these assumptions in a   composite
sense,  well-water concentrations could increase by at least an
order of magnitude.

Based  on  the results from the saturated zone simulations  and
the    preceeding   discussion,   there   are   a   number   of
recommendations  which  can  be made that will tend  to  reduce
expected well water concentrations.

    1)   Avoid application of pesticide up gradient of wells
         that  intersect major solution cavities in areas where
         the  aquifer  is unconfined or where  direct  linkages
         exist.

    2)   Avoid application near wells in the surficial aquifer
         that   have   high  pumping  rates  and  are   closely
         surrounded by citrus groves.

    3)   Avoid using shallow wells where the localized induced
         gradient is very large.

    4)   A well tapping a confined aquifer will be safer than
         unconfined aquifer.
                              45

-------
5)    Deep  wells  that draw water from a large  section  of
aquifer  will  have lower concentrations than similar  wells
that are shallower.
                          46

-------
                           SECTION 2
              FLORIDA CITRUS GROWING ENVIRONMENT
Site-specific  information about the citrus growing region  was
gathered  to delineate a number of scenarios among which it was
felt   the   fate   and  transport  of   the   chemical   would
significantly  differ.   The information gathering was  divided
into  two  topical areas:  1) the surface and unsaturated  zone
and  2)  the saturated zone.  These two zones are discussed  in
the  following  subsections in terms of the important  physical
characteristics  that  distinguish  unique  scenarios  for  the
purposes of modeling.
2.1  SURFACE AND UNSATURATED ZONE
Citrus  is  grown  throughout  most  of  central  and  southern
Florida.   The  map  in Figure 2.1 shows the  counties  of  the
state   ranked   according  to  the  acreage  of  bearing   and
non-bearing  grapefruit and oranges grown.  The top ten  ranked
counties  are highlighted.  The preponderance of fruit grown is
in  the  "ridge" area, a region of rolling sand  hills  running
approximately  north-south in the center of the state, and  the
Indian  River  area,  the  smaller shaded region  on  the  east
                              47

-------
     ALABAMA
                                                               Jacksonville
                                                                     Atlantic Ocean
                   1-PolJc
                   2-Laka
                   3-St. Lucle
                   4-Zndian River
                   5-Orange
                   6-Hardee
                   7-Martin..
                   8-H111eborough
                   9-Highlands
                  10-De Soto
                  11-Pasco
                  12-Hendry
                  13-Osceola
                  14-Brevard
                  15-Manatee
                  16-Palm Beach
                  17-Marion
                  18-Volusia
                  19-Collier
                  20-Okeechobee
                  21-Lee
                  22-Hernando
                  23-Seminole
                  24-Charlotte
                  25-Glades
                  26-Pinellas
                  27-Putnam
                  28-Broward
                  29-Sumter
                  30-Sarasota
                  31-Citrus
                  32-Plagler
                  33-Alachua
                  34-St. Johns
Figure  2.1   Florida  counties ranked by citrus acreage.
                                   48

-------
coast.   These  are rankings as of January 1, 1982.   A  severe
freeze  occurring in December 1983 killed substantial  acreages
in  the  northern portion of the state.  Post  freeze  acreages
were  not  available.  Table 2.1 gives the acreage  by  county.
Over  half the citrus is grown in the top five ranked counties;
over 75% is grown in the top ten ranked counties.
2.1.1  Climate
The  driving  force  behind  pesticide transport  in  soils  is
primarily  water  movement downward in the profile; whether  it
occurs  from natural rainfall or irrigation.  Rainfall is  high
in  the  Florida citrus growing region, averaging from  132  to
162  cm/yr  (52  to 64 in./yr, .Figure  2.2).   Pan  evaporation
ranges from 117 to 132 cm/yr  (46 to 52 in./yr , Figure 2.3).

However,  65% of the rainfall on the average occurs during  the
months   of   June,   July,  August  and   September.    Evapo-
transpiration   demands   require  that   irrigation  be   used
extensively  in  March,  April and May  to  supplement  natural
rainfall  and  50% of the irrigation water applied  is 'applied
during  these months (Harrison, 1984, personal  communication).
Table  2.2  shows  the monthly distributions  of  rainfall  and
evapotranspiration  for Lake Alfred (located in the ridge area)
and  Fort  Pierce  (flatwoods area)(see  Section  2.1.2).   The
differences  indicate  that,  on the  average,  water  deficits
occur  in  the  Lake  Alfred area in  fall,  winter  and  early
spring,  while  at  Fort Pierce, deficits  generally  occur  in
spring.

Temperatures  in  the area are normally mild,  however  freezes
may  occur.  These conditions may affect chemical transport  as
irrigation  systems are often used to apply large quantities of
                              49

-------
    TABLE  2.1   SUMMARY OF  BEARING AND NONBEARING
                 CITRUS  (ORANGES AND GRAPEFRUIT)
                 BY COUNTY AS OF JANUARY 1,  1982
County
Acreage
County
Acreage
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Polk
Lake
St. Lucie
Indian River
Orange
Hardee
Martin
Hillsborough
Highlands
De Soto
Pasco
Hendry
Osceola
Brevard
Manatee
Palm Beach
Marion
125331
104307
71057
60387
42727
42136
38714
34807
34210
32902
31636
29962
16550
14875
12980
12275
10840
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Volusia
Collier
Okeechobee
Lee
Hernando
Seminole
Charlotte
Glades
Pinellas
Putnam
Broward
Sumter
Sarasota
Citrus
Flagler
Alachua
St. Johns
8786
7561
. 6668
6254
5819
5790
5757
3933
2245
2129
1661
1560
1258
1144
116
96
95
                                        TOTAL   776,568
                                                      11
Source:  Florida Citrus Mutual,  1984.
Acreage for 1982-83  crop season was 776,803 with 71,053 additional
acres  in specialty fruit according to the Florida Dept. of
Agriculture and Consumer Services, -1984.
                          50

-------
              ALABAMA
                                                 Jacksonville
                                                       Atlantic Ocean
                                                        52
Figure 2.2   Annual precipitation  in inches (1 inch =  2.54  cm).
             Source:  U.S.  Dept.  Commerce, 1972.
                                 51

-------
         ALABAMA
                                            Jacksonville
              Gulf of Mexico
                                         ..'*
Figure 2.3   Annual lake evaporation in inches  (1  inch  =
             2.54 cm).  Source:   Kohler et al., 1959.
                            52

-------
TABLE 2.2  MEAN MONTHLY RAINFALL AND EVAPOTRANSPIRATION AT TWO
           CITRUS GROWING LOCATIONS IN FLORIDA  (cm)
                 LAKE ALFRED
FORT PIERCE
Month
J
F
M
A
M
J
J
A
S
0
N
D
TOTAL
Rainfall1
4.9
6.4
7.9
8.3
8.4
20.6
19.3
19.3
9.9
4.7
4.7
4.6
133.1
ET2 .
5.3
5.8
5.8
9.9
12.7
15.0
14.5
12.4
10.2
7.4
7.4
5.8
121.1
Diff
-0.4
0.4
0.0
-1.6
-4.3
5.5
4.8
6.9
-0.3
-2.7
-1.2
-1.2

Rainfall
5.4
5.0
7.0
9.8
13.2
19.0
17.1
16.9
23.4
21.2
6.8
5.2
150.0
ET2
5.3
6.6
9.1
11.4
13.5
11.2
12.4
12.2
10.2
9.1
6.8
5.3
113.3
Diff
0.1
-1.6
-2.1
-1.6
-0.3
7.8
4.7
4.7
13.2
12.1
0.0
-0.1

 Rainfall Depths are approximate.
 Source:  U.S. Dept of Commerce, 1968
 "Source:  Harrison, 1984.
                               53

-------
water  for  freeze protection.  The annual mean number of  days
having  temperatures  of 32 deg. F (0 deg.C) or less are  shown
in  Figure  2.4.   Citrus in the upper portion of  the  growing
area  may  experience close to 5 days of freezing  temperatures
annually.   Almost all citrus could expect at least 2  freezing
days annually.
2.1.2  Soils
There  are  four orders of soils in Florida on which citrus  is
extensively  grown.  These are the entisols, ultisols, alfisols
and  spodosols.  The approximate locations of these groups  are
shown  on the accompanying map (Figure 2.5).  The entisols tend
to  occur in the central "ridge" area with the major occurrence
of  ultisols  being in the northwestern portion of  this  area.
The  spodosols  occur  along  the middle  eastern  and  western
coasts  of  the  peninsula, adjacent to the ridge, in  what  is
known  as the  "flatwoods"  area.  Two major areas of  alfisols
occur;  one running north-south which divides the eastern coast
flatwoods  area and the other immediately east and south of the
ridge  area.   Naturally, there are isolated pockets  of  these
soil types occurring everywhere.

The  entisols  are  soils which are loose, incoherent  and  are
developed   from   thick   beds  of  marine   sediments.    The
characteristic  of  these soils is a distinct lack  of  profile
development,  the only exception being slightly higher  organic
matter  in the surface soil.  They are strongly acid, are  well
drained   to  excessively  drained,  and  their  water  holding
capacities are low  (Anderson, 1981).

The  ultisols  are  highly weathered soils with  sandy  surface
horizons  but  loamy to clayey subsurface horizons.   In  almost
                              54

-------
10
                                                  Atlantic Ocean

                                                    2
 Figure 2.4  Mean annual  number of days having a minimum
             temperature  of  32°F (0°C)  or below.
             Source:  U.S. Dept of Commerce,  1968.
                             55

-------
     ALABAMA
                     Mostly Ultisols

                     Mostly Entisols,
                      some ultisols
                     Mostly Spodosols

                     Mostly Alfisols,
                      some spodosols
                                                   Jacksonville
                                                       Atlantic Ocean
Figure 2.5   Location and extent of  soil orders on which
              citrus  is grown.
              Source:   Caldwell and Johnson,  1982.
                           56

-------
all  other  aspects  they  are very similar  to  the  entisols.
Their   water   holding  capacity  and  CEC  (cation   exchange
capacity)  rises substantially in the clay accumulation  layer,
which  is  typically  at a depth of 100 to  200  cm  (Anderson,
1981).

The  spodosols  occuring along the eastern and  western  coasts
have  an illuvial (accumulation) horizon located normally 40 to
120   cm  below  the  surface.   These  soils  are   frequently
saturated  unless  artificially drained.  Associated with  this
may  be  high  accumulations of organic matter in  the  surface
horizon  with  a  hard  pan in the  eluvial  (leached)  horizon
(Brady, 1974).

Alfisols  are  moist mineral soils which are weathered but  not
to  the  extent of the spodosols or ultisols.  They are  higher
in  base  saturation  with  higher pH  than  the  spodosols  or
ultisols.   They  have an illuvial horizon in which clays  have
accumulated.

It  was  felt  that the transport  characteristics  might  vary
substantially  among  these soil orders.  To investigate  their
transport  properties,  soils which are commonly  planted  with
citrus   (Tucker,  1978) were subdivided into  their  respective
orders.   Then,  occurrences  of samples of  these  soils  were
located  in  soil  characterization  data  of  Calhoun  et  al.
(1974),  Carlisle  et  al.  (1978) and Carlisle et  al.  (1981).
Samples  having  significant amounts of missing information  in
the  profile  were  dropped.   The remaining  data  on  organic
carbon,  saturated  hydraulic  conductivity,  pH,  texture  and
water  holding capacities were entered into a computerized data
base.   Table  2.3 shows the soils and sample numbers used  for
characterization.    There  were  a  total  of  19  samples  of
entisols, 9 of ultisols, 20 of spodosols, and 10 of alfisols.
                              57

-------
TABLE  2.3 SOIL SAMPLE SELECTED FOR SOIL CHARACTERIZATION
           ANALYSIS
 Soil
         Number of
Order     Samples
Astatula   Entisol


Candler    Entisol

St. Lucie  Entisol
Paola
           Entisol
Wabasso

Pineda

Felda

Riviera
                         2

                         5
Arredondo  Ultisol       6


Apopka     Ultisol       3

Immokalee  Spodosol      2

Myakka     Spodosol      9



Oldsmar    Spodosol      5
           Spodosol      4

           Alfisol       3

           Alfisol       2

           Alfisol       5
                                         Sample Numbers
                       S27-22, S49-8, S64-4, S64-5,
                       S5-1, S5-24, S5-26

                       S27-6, S49-5

                       S50-20, S56-10, S5-27,
                       S49-1, S49-7

                       S27-7, S50-21, S64-6, S5-2,
                       S49-6

                       S27-11, S35-1, S35-2,
                       S42-119, Sl-6, Sl-84

                       S64-12, S35-3, S53-4

                       S50-17, S64-8

                       S27-14, S27-16, S49-10,
                       S56-6, S64-16, S5-6, S5-7,
                       S50-1, EXP. S50-1

                       S49-30, S50-33, S56-7, 556-12,
                       S56-2

                       S50-13, S64-24, S56-1, S51-2

                       S5-17, S50-10, S43-5

                       S50-2, S51-7

                       S49-15, S56-18, S5-13, S43-18,
                       S56-21
 Sample numbers  correspond to those used in Calhoun et al. ,
 1974; Carlisle  et al.,  1978; and Carlisle et al., 1981.
                                58

-------
To  analyze  profile  characteristics, each soil  was  assigned
parameter  values  for  depths  from  1  to  300  cm  at  1  cm
increments.   This  was  done  by  interpolating  between  data
points  and assigning values for increments above or below  the
range  of  measurements  to the shallowest and  deepest  values
found,  respectively.  Arithmetic means and standard deviations
for  each parameter were then computed at each 1 cm  increment.
The  results  were then plotted.  (For detailed plots  of  mean
parameter value with standard deviations see Appendix B).

Figure  2.6  shows  the  means  of  the  field  capacity  water
contents  (6 fc , % by volume) for five groupings of soils.   In
PRZM,  9fc  affects the rate of percolation of water, and hence,
of  the  chemical, through the profile.   The  Astatula-Candler
and  St.  Lucie-Paola  groupings were left separate to  see  if
differences  could  be discerned between them since the  latter
are  entisols  occuring  primarily  in  the  "flatwoods."   The
curves  show that  0 fc is nearly identical for all the entisols,
being  roughly 5-7% at the surface, then dropping rapidly,  and
asymtotically  approaching  an  average value of  2-3%  in  the
lower  profile.  The spodosols and alfisols, on the other hand,
show  a  higher  9 f in  the surface, 14 to  15%,  which  drops
rapidly  to 5-6% at a depth of 20-60 cm and sharply rises again
to  25-26% at depths greater than a meter.  The ultisols show  a
similar  trend, with slightly lower  6 fc at the surface and  the
high   0£c layer  occurring  lower in the profile  than  in  the
spodosols or alfisols.

Interestingly,   the   mean  values   for  saturated   hydraulic
conductivity  (Kg, cm/hr)  (Figure 2.7)  show patterns which  are
virtually  the  mirror  image of the curves of  6 f c  vs.  depth.
Again  the  entisols are the deviant group for this  parameter.
Of  particular interest is the fact that for the  soils  having
the  lowest  mean  hydraulic conductivities  (alfisols),  water
transmission  rates  are   still  in excess of  1  cm/hr  or  24

                              59

-------
                   PARAMETER 0f(,, IN PERCENT

             5.0      10.0      15.0      20.0     25.0
                               Explanation


            0 Candler  & Astatula (ridge entisols)
            4 Arredondo & Apopka (ultisols)
            • Oldsmar, Inunokalee, Wabasso, Myakka (spodosols)
            D Felda, Riviera, Pineda (alfisols)
            • St. Lucie, Paola (flatwood entisols)
Figure 2.6  Mean values of field capacity water content vs
            depth from soil characterization analysis.
                           60

-------
             10
            PARAMETER KS/  IN  cm/hr

        20      30      40      50
70
        Not
                              Explanation

            O Candler & Astatula (ridge entisols)
            4 Arredondo, Apopka (ultisols)
            • Oldsmar, Immokalee, Wabasso, Myakka (spodosols)
            D Felda, Riviera,  Pineda  (alfisols)
       shown  st* Lucie & Paola  entisols  are  60-90  cm/hr
Figure 2.7
Mean values for .saturated hydraulic conductivity
versus depth from soil characterization analysis,
                              61

-------
cm/day.   This has important implications for unsaturated  zone
modeling  in  that water and fluxes may not be limited by  soil
transmission  rates  but  by  water  table  conditions  at  the
interface of the unsaturated and saturated zones.

Figure  2.8  shows the mean soil pH versus depth for  the  five
soil  groupings.  The pH profiles are closely grouped within  a
range  of roughly 5.0 to  6.0 except for the alfisols which are
typified  by  pH values of 6.0 to 7.0.  While these  pH  values
are  not much higher, base-catalyzed aldicarb hydrolysis  rates
may   be   substantially  higher  in  these  soils.    Standard
deviations  are generally +0.5 in the entisols and ultisols and
+1.0 in the spodosols and alfisols (see Appendix B).

Figure  2.9 shows soil organic carbon versus depth for the five
soil  groupings.   The quantity of organic carbon in  the  soil
affects  the adsorption of the chemical to soil materials.  For
this  parameter,  all soils show a nearly  monotonic  reduction
with  depth  except  for the spodosols  which  show  subsurface
accumulation, peaking at about 90 cm.

Thus   we   see   that  three  of  these   soil   orders   have
distinguishing  characteristics which in turn have implications
for  aldicarb trknsport; the entisols with their very low water
holding  capacities, the alfisols with their higher pH and  the
spodosols  with  their  accumulation layer of  organic  carbon.
The  ultisols  do not distinguish themselves in  any  category,
which  by  default categorizes them in a range of  intermediate
properties.

The  entisols  stand  out  as  the  group  for  which  aldicarb
transport  is  potentially  highest, having low  water  holding
capacity,  low  pH  and low organic carbon.   Therefore,  these
soils  should  show  the lowest degradation  rates  and  lowest
chemical  adsorption  although the low water  holding  capacity

                             62

-------
      3.0
E
u
w
Q
4.0
PARAMETER pH



5.0      6.0
7.0
                            Explanation


             0 Candler & Astatula (ridge entisols)

             • Arredondo & Apopka (ultisols)

             • Oldsmar, Immokalee, Wabasso, Myakka  (spodosols)

             D Feld, Riviera, Pineda (alfisols)

             • St. Lucie, Paola (flatwood entisols}
    Figure 2.8   Mean values of soil pH versus depth from the
                 soil  characterization analysis.
                             63

-------
             PARAMETER ORGANIC CARBON, IN PERCENT

              0.5       1.0       1.5       2.0
                                        2.5
                          Explanation

       0 Candler & Astatula (ridge entisols)
       f Arredondo & Apopka (ultisols)
       •Oldsmar, Immokalee, Wabasso, Myakka (spodosols)
       DFeld, Riviera, Pineda (alfisols)
       • St. Lucie, Paola (flatwood entisols)
Figure 2.9
Mean values of organic Carbon versus depth
from soil characterization analysis.
                          64

-------
may somewhat counteract the low organic carbon contents.
2.1.3  Irrigation
Irrigation  is practiced in Florida citrus to boost yields, and
to  protect against droughts and freezes.  It is essential  for
production  in the coastal flatwoods.  Approximately 62% of the
citrus  is  irrigated  by some method (Stanley et  al.,  1980).
Irrigation  water  provides an additional opportunity  for  the
leaching of the chemical.

Although  a  variety  of scientific methods  are  available  to
schedule  the application of water, most growers, by virtue  of
intuition  or other indicators,  normally initiate  irrigations
at  50%  available  water depletion  (Harrison,  1984;  Personal
Communication).

There  are,  at  present,  five  major  methods  of  irrigation
utilized  in  Florida citrus.  These are 1) permanent  overhead
sprinklers,   2)  traveling volume guns,  3) low volume  spray,
4)  flood, and  5) drip.  Stanley et al. (1980) subdivided  the
citrus  growing  region into three areas; the upper ridge,  the
lower  ridge, and the flatwoods.   The upper ridge  corresponds
to  the  area  covered  by entisols  lying  above  the  southern
boundaries  of Lake and Orange counties including Lake, Marion,
Orange  and Pasco counties.  The lower ridge corresponds to the
entisol  area below this division made up primarily of  Hardee,
Highlands,  Hillsborough  and  Polk  counties.   The  flatwoods
include   areas  on  both  the  eastern  and  western   coasts,
corresponding  to  the  spodosols and  alfisols  and  including
primarily  Brevard, Desoto, Hendry,  Indian River, Martin,  Palm
Beach and St. Lucie counties.
                              65

-------
Table   2.4  shows  the  breakdown  of  these  five  and  other
irrigation   types  by  subarea.   According  to  this  survey,
traveling  guns  and permanent overhead guns were  the  favored
methods  in the upper ridge, while these plus low volume  spray
were  favored in the flatwoods along with a high percentage  of
portable  guns.   Harrison  (1984, personal communication)   has
indicated  that  the usage of traveling guns has diminished  by
50%,  portable guns by 90%, seepage by 20%, perforated  pipe by
80%  (there  is practically no irrigation done by  this  method
currently),  and  permanent overhead by 20%.  There has been  a
corresponding  increase  in the number of drip and  low  volume
spray  systems  since  1980.  The characteristics  of  each  of
these  methods and the use of irrigation for freeze  protection
will  be  discussed individually in the following sections.   A
summary  of system management characteristics is given in Table
2.5.

2.1.3.1  Permanent Overhead Sprinklers—

These  systems consist of rotating sprinklers which are located
at  intervals  through the orchard on high risers which  .extend
several  feet  above the tree canopy.  The entire'area  of  the
grove  is wetted during irrigation events.  Water is applied at
rates  of  approximately  0.5 cm/hr  (Harrison,  1984,  Personal
Communication).   Four  to  five  (4-5) cm of water  are  applied
per  event at  5 to 7 day intervals.  An average of about 38  cm
of  water  was applied annually by permanent  overhead  systems
monitored  in  the state from 1970 to 1980  (Duerr and  Trommer,
1982).

2.1.3.2  High  Volume Guns—

Self   propelled   and  portable  guns  have  the  same   basic
characteristics.   They both discharge a large volume of water,
.0079-.076  m/s   (125 to 1200 gpm) under high pressure   from   a

                             66

-------
TABLE  2.4  CITRUS IRRIGATION SYSTEMS WITH ACREAGES FOR EACH OF THREE AREAS IN
            THE  STATE
AREA
I.
Upper
Ridge
II.
Lower
Ridge
III.
Indian
River fi
Flat-
woods
TOTAL
Traveling
Guns
Acres
31,044
33,968
5,576
70,588
%
37.0
21.6
1.9
13.3
Portable
Guns
Acres
7,7375

105,740
113,115
%


36.6
21.4
Seepage
Acres


147.169
L47.169
t


51 JO
51. (
Perforated
Pipe
Acres
4,526
9,532
260
14,318
%
5.4
6.1
0.1
2.7
Permanent-
Overhead
Acres
33.24C
80,421
12,510
L26.171
1
39.7
51.2
4.3
23. e
Drip
Acres
4,735
9,956
7,656
22,347
%
5.6
6.3
2.6
4.2
Low Volume
Spray
Acres
2,891
23,164
9,996
\
36,051
%
3.4
14.8
3.5
6.8
Total
Irrigation
Acres
83,811
157.041
288,759
529,759
 Sourcet  Stanley, et. al., 1980

-------
     TABLE  2.5   SUMMARY  OF MANAGEMENT CHARACTERISTICS OF IRRIGATION SYSTEMS
  Type
     Event
Application Rate
 in Wetted Area
     (cm/hr)
Irrigation Frequency   Wetted Area

      (days)               (%)
              Approximate
                Annual
              Application

                 (cm)
  Permanent
  Overhead
  High Volume
  Gun
CO
  Seepage
  Drip
      0.5
      0.5
   0.06-0.122
      0.7 =
        5-7



        8-10



       10-13


      0.3-1
100
100
100
38



31



24


36
  Microjet
      0.62
        2-3
 171
22
    Assumes a unit area of 58 m2 per tree  (25x25  ft  spacing)
    Calculated

-------
single  nozzle.   The  water  is shot high  into  the  air  and
impacts  the ground and tree canopy in much the same way that a
heavy  rainfall would.  The frequency and depth of  irrigations
are  much  the  same as for permanent  overhead  systems.   The
average   amount   of  water  applied  per  year  by  guns   is
approximately 31 cm (Duerr and Trommer, 1982).

2.1.3.3  Seepage (Flood)—

Seepage  irrigation  is used only in the flatwoods areas  where
there   is  either  a  shallow  water  table  or  a  relatively
impermeable   layer  close  to  the  soil  surface.   Water  is
introduced  into  furrows adjacent to the  beds.  Depending  on
how  high the water is allowed to rise in the furrow, this type
of  irrigation  is either called furrow flood or  crown  flood.
With  crown flooding, water is brought to the level of the  top
of  the  bed,  whereas with furrow flooding it  is  brought  to
roughly  one half the furrow height, relying on capillary  rise
to  moisten soil in the beds above this height.  The amount  of
irrigation  water  reaching the root zone is a function of  the
permeability  of  the soil and the residence time of the  water
in  the furrows.  Calvert et al. (1967) showed that it took  13
days  on  a  Felda  soil and 10 days on an  Immokalee  soil  to
complete  a  single  wetting   and drying  cycle  using  furrow
irrigation.   Capillary water rise  above the free water  table
was  measured at 15.2 cm (6 in.) in an Immokalee soil and  30.5
cm  (12  in.) in a Felda soil.  Based on monitoring  data  from
1970  through 1978, it was determined that about 24 cm of water
on  the  average  is applied annually by  furrow  type  systems
(Duerr and Trommer, 1982).

2.1.3.4  Drip  (Trickle)—

Drip  irrigation is a highly efficient method of applying water
directly  to  the crop root zone.   A low pressure  water  line

                              69

-------
running  down  each  row of trees is tapped by  emitters  which
drip  water  in an approximately 1 sq. m area around the  tree.
Emitters  apply water at a rate of about 1.07 cm/s (0.017  gpm)
(there  are normally two emitters per tree) and irrigations are
needed  two  to  three times per day during peak  use  periods.
Duerr   and   Trommer  (1982)  monitored  the  average   annual
application  by  drip systems to be 36 cm.  This is on  a  unit
area  (e.g.,  per acre) basis, however, and application in  the
wetted area would be much higher.

2.1.3.5  Low Volume Spray—

Low  volume  spray  systems  are similar to  drip  systems  but
operate  at higher pressures and wet a larger area around  each
tree.   The  typical  wetted area is approximately  10  sq.  m.
These  systems  apply water at an average flux of roughly  15.8
cm/s  (0.25 gpm).  Irrigations are needed two to three times  a
week.   Duerr  and Trommer (1982) measured the  annual  average
application  depth of low volume spray systems to be 22 cm on a
total  areal basis.  Application rates are higher in the wetted
zone.

2.1.3.6  Irrigation For Freeze Protection—

Most  of the methods of irrigation previously mentioned can  be
used  for freeze protection in citrus.  There are two  measures
which  can  be taken, one which is practiced before the  freeze
and one practiced during the freeze periods.

The  idea  behind  pre-freeze  irrigation is  to  increase  the
thermal  conductivity and heat capacity of the soil.  Following
a  freeze  warning,  0.6 to 1.3 cm of water is  applied.   This
allows  the soil to store heat which will be released into  the
grove during lower temperature periods.
                              70

-------
w
OT

Pi

W
CO
w
ffi
o
          w
00
f


03
    I

   00
*r
 t

o
          03
           •

          O
                                      Illuvial Layer
            %ssss8%s^g^ss&s^^
                            5-8
                                        25
                                                     5-8
FEET
                      1.2 1.5-2.4
                                        7.6
                                                   1.5-2.4  1.2 METERS
Figure 2.10  Typical configuration for Bedded citrus in the Flatwoods Area.

-------
Irrigation  water  applied  during  the  freeze  provides  both
sensible  heat contained in the water and latent heat of fusion
as  that  water becomes ice.  Enough water must be  applied  so
that  the sensible plus the latent heat of fusion added exceeds
the  convective, radiation and evaporative heat losses from the
plant.   The convective and radiation heat losses are of  equal
magnitude  for  a  wet  or  dry leaf.   Heat  consumed  due  to
evaporation  of  water, however, exceeds the heat liberated  by
fusion  by a factor of about 7.5.  Therefore, 7.5 times as much
water  must  be frozen as is evaporated in order to  raise  the
temperature.   Because  high  winds increase  evaporative  heat
losses,  this  method  is  safe to use  only  during  radiation
freezes.   In advective freeze, more harm than good can be done
(Harrison  et  al.,  1974).  Irrigation only by the  drip,  low
volume  spray  or flood methods are used during  freezes.   The
use  of  overhead systems results in severe breakage  of  limbs
due  to ice build up and is feasible only when trees are  young
(i.e.,  in nurseries).  Where low volume sprays are used  water
is  applied at the sprinkler capacity until the temperature  in
the  grove  reaches  32  deg. F  (0  deg.  C)  (Harrison,  1984,
personal communication).
2.1.4  Cultural Practices
2.1.4.1  Bedding—

The  chief  cultural practice which distinguishes ridge  citrus
from  flatwoods citrus is that of forming beds for trees in the
flatwoods areas.

A  typical  configuration for double bedded citrus is shown  in
Figure  2.10.  Other configurations may have as few as one  row
per  bed or as many as eight.  Tree spacing is typically 7.6m

                             72

-------
(25  ft) between rows and varies from 3.8m (12.5 ft) to 7.6  m
(25  ft)  down  each  row.  The land is  formed  by  excavating
furrows  between each bed and using this soil to create  raised
beds  in  the middles.  This provides additional rooting  depth
for  trees  and  provides  a method of  readily  irrigating  or
draining  beds.  The bed heights are typically 0.3 to 0.8 m (12
to   30   inches).   At  times  bedded  citrus  may   also   be
artificially  drained  by tile or corrogated plastic pipe.   If
so,  these  drains are placed three to four feet below the  top
of  the  bed,  either in the bed centerline or in  the  furrows
just  above  the illuvial layer on a 0.001 to 0.002 m/m  slope.
This  practice  of draining is probably used on less than 5  to
10%    of   the   flatwoods   area   (Woods,   1984,   personal
communication).

In  the  ridge areas beds are unnecessary due to the  extremely
good   drainage  characteristics  of  the  soils.   Trees   are
typically  planted  on 7.6 by 7.6 m (25 by 25 ft)  spacing.  In
the  ridge areas trees normally attain greater canopy dimension
than in the flatwoods.

2.1.4.2  Control of Volunteer Vegetation—

Herbicides  are  generally applied to 1.2 to 1.5 m  (4 to 5  ft)
swaths  on each side of the trees in bedded areas to  eliminate
weeds.   This  has the advantage of reducing  transpiration  by
eliminating  unwanted  vegetation  in the root  zone  and  also
eliminating   weeds   which  might  interfere  with   sprinkler
patterns  of  low volume jets.  On the remainder of  the  area,
natural  vegetation  or sod is used to reduce erosion  of  beds
and  furrows.  This is either mechanically or chemically mowed.
In  the  ridge areas weeds are controlled using  herbicides  or
discing  to reduce transpiration by volunteer vegetation, clean
sprinkler  patterns  and  to promote drainage of cold  air  off
hillsides during freezing periods.

                             73

-------
2.1.4.3  Pest Control—

Common  to  both areas is the occurrence of  nematodes.   These
pests  occur in the greatest number from 0.3 to 1.8 m below the
surface  when  there  is the greatest number of  feeder  roots.
Usually  the  top  15 cm is too hot and dry  to  support  them.
(Anderson,  1981).   Mites  are also problematic  in  the  tree
canopy.   Both  of these pests can be controlled by the use  of
aldicarb.   The chemical is normally chiseled in to a depth  of
5  cm,  in  1  to 1.5 m bands on each side  of  the  tree  rows
centered under the drip line.

2.1.4.4  Fertilization—

Groves  are  normally limed to maintain pH between 6.0 and  7.0
in  the  root zone.  As mentioned previously, high soil pH  may
increase aldicarb degradation due to hydrolysis.
2.1.5  Thickness of Unsaturated Zone
The  thickness of the unsaturated zone is a critical factor . in
the  transport  of  aldicarb  from  the  soil  surface  to  the
saturated  zone.   Where the unsaturated zone is thin  and  the
water  table is close to the land surface, aldicarb will  enter
the  aquifer  much  more quickly and have less  opportunity  to
degrade  than  where the unsaturated zone is thicker.   In  the
citrus   growing   areas  of  Florida  the  thickness  of   the
unsaturated zone depends on the topographic region.

The  central peninsula area falls into two physiographic areas;
the  central highlands and the coastal lowlands (Cooke,  1945).
The  central  highlands is composed of a series of long  narrow

                              74

-------
ridges  30.5 to 61 m (100 to 200+ ft above sea level) that  run
parallel  to  the  axis of the  peninsula.   Surrounding  these
ridges  is a broad upland area of lower elevation 15.3 - 30.5 ra
(50-100  ft) and much less local relief.  The coastal  lowlands
(also  known  as the flatwoods) flank the central highlands  on
the  Atlantic and Gulf of Mexico coasts.  This area,  generally
below  15 m (50 ft) in elevation, consists chiefly of old beach
ridges,   terraces,  lagoons,  and  swamps.   The   generalized
topographic  map   (Figure  2.11) clearly shows  these  general
features.

The  thickness  of  the unsaturated zone  closely  follows  the
topograpic  trends.  In the poorly drained lowlands, the  water
table  is  usually  within  1.5 m (5 ft) of  the  land  surface
(Healy,  1974).   In  these  areas the citrus  groves  must  be
drained  for   successful  cultivation.   The  water  table  is
controlled  by  a  system of canals, so the  thickness  of  the
unsaturated zone is dependent on man's activities.

In  the central highlands, the depth to the water table is much
more  variable.   Near lakes the unsaturated zone can  be  very
thin.  Beneath the high ridges it can be as much as 30.5 m (100
ft)  thick.  Generally, the unsaturated zone averages 3 to 9  m
(10  to  30  ft)  in  thickness.   Figure  2.12  is  a  typical
cross-section   through  the  ridge  area  near  Lake  Hamilton
(central  highlands)  showing the variability in  thickness  of
the unsaturated zone.
2.1.6  Delineation of Unsaturated Zone Scenarios
Based   on  the  preceeding  information  on  climate,   soils,
irrigation   methods,   cultural   practices,  and   depth   to
saturation,  six  subareas  of the citrus growing  region  were
                              75

-------
           ALABAMA
                                                        Jacksonville
                   Explanation

         Elevation in feet above sea level

                      0-50

                      50 - 100

                      100 - 150

                      Greater than 150
                                                             Atlantic Ocean
Figure 2.11   Generalized 'topography of  the Central Peninsula,
               Florida.   Source:  USGS,  1:250,000  scale topo-
               graphic  sheets  of  Florida.
                                   76

-------
                                                              01
w
                       6    8   10   12   14    16   18   20

                        DISTANCE,  IN  THOUSANDS  OF  FEET
22   24
   Figure 2.12  Typical cross-section  through  a  ridge citrus  area
                showing relative  thickness  of  unsaturated  zone.
                Source: ; Anderso'n and  Simonds, 1982.

-------
identified.    It   was  felt  that  these  regions   represent
distinctly  unique situations with respect to aldicarb leaching
in  the unsaturated zone.  These areas are shown on the map  in
Figure  2.13.   The boundaries of these areas  are  approximate
and  do not necessarily correspond to any political  boundaries
or natural landforms found within the state.

Subareas  1  to  4  correspond to  locations  having  entisols,
ultisols,  spodosols,  and  alfisols,  respectively,  with  two
subareas  of wetter spodosols and alfisols broken out (subareas
5  and 6).  Areas 5 and 6 receive roughly 10% or more  rainfall
annually than areas 1 to 4.

Table   2.6  shows  typical  characteristics  of  these  areas.
Rainfall  differences  are not large across the  entire  citrus
growing  region,  but higher annual depths do tend to occur  in
areas  5 and 6.  These differences will be mitigated in certain
cases  by  the  addition  of irrigation  water.   The  rainfall
depths  shown  in  Table 2.6 are the means for  the  period  of
record   (1965-1978) at Lake Alfred (areas 1-4) and Belle  Glade
(areas  5  and  6)  actually used in  model  simulations.   The
potential  evapotranspiration  gradient runs from northeast  to
southwest  with upper area 3 (east coast) tending to be  lowest
and  area  3   (west  coast) and areas 5 and  6  tending  to  be
highest.   Actual values used in the simulation were also  from
the Lake Alfred and Belle Glade meteorological records.

In  each of these areas, each of four irrigation practices  are
utilized.   Areas  1 and 2 generally receive no  irrigation  or
irrigation  by  drip,  low volume spray,  or  overhead  systems
while  areas   3,  4, 5 and 6 will tend to be irrigated  by  the
drip,   low  volume  spray,  overhead  or  flood  methods.   If
aldicarb  is   applied on treated bands centered underneath  the
drip  line  in a mature orchard, however, the wetted area of  a
drip  irrigation  system will not intersect the  treated  band.

                              78

-------
                                                       Jacksonville
                                                            Atlantic Ocean
Figure  2.13  Subareas  delineated  for unsaturated zone modeling.
                                   79

-------
 TABLE 2.6   CHARACTERISTICS OF AREAS SELECTED FOR UNSATURATED
            ZONE MODELING


         Annual       Annual     Soil1    Irrigation2 Depth to
      Precipitation    ETp       Order      Method    Saturated
Area	(cm)'  	(cm)   	Zone (m)
1
2
3
4
' 5
6
121
121
121
121
140
140
117-127
117-122
114-137
122-132
127-132
127-132
E
U
S
A
S
A
L./O,N
L,0,N
L,0
L ,0
L.,0
L ,0
2.7, 9.
1.8, 9.
1.2
1.2
1.2
1.2
0
0




 E = Entisols,  U = Ultisols, S = Spodosols, A = Alfisols
2 L = Low volume spray
 0 = Overhead  (Permanent Sprinklers or Guns)
 N = None
                              80

-------
Therefore,   this   irrigation  method  was   eliminated   from
consideration.   Neither were the flood methods considered.  To
adequately  model the movement of water under these  irrigation
systems  would  require a two-dimensional model.   Even  though
PRZM  was  modified, (as discussed in Appendix A), to  simulate
lateral  drainage  of water from bedded citrus, the  simulation
of   the   addition  of  water  would  necessarily  be   overly
simplistic  in  a  one-dimensional framework.  It  was  decided
that  the flood irrigation method would be eliminated in  favor
of presenting results based on questionable simulations.

Consideration   of   freeze  protection  irrigation  was   also
eliminated.   The  duration  of  the application  of  water  is
highly   variable  and  dependent  upon  the  duration  of  the
freezing  period.   These periods normally occur at night  with
temperatures  rising  above freezing occurring during the  day.
Therefore, mean daily temperatures are rarely below freezing.

PRZM  requires mean daily temperatures as an input in order  to
calculate  potential  evapotranspiration if pan evaporation  is
missing.   Based only on mean daily temperature, PRZM might not
be  able  to  detect  the occurrence  of  a  freeze  protection
irrigation   event.    Even  if  maximum  and   minimum   daily
temperature  were  input, the duration of the  freezing  period
would  still  be  unknown.  Proper simulation of  these  events
would  require  at least hourly temperature inputs,  which  are
simply  not  available,  and  wind speed to  determine  if  the
freeze  is radiative or convective.  In the end it was  decided
that  due  to the complexity of the processes and  data  inputs
required,  meaningful  simulations could not be done, with  the
time and available budget.

Areas  1  and  2 show the greatest variability  in  unsaturated
zone  thickness.    It was decided that at a minimum two  depths
should  be  simulated.  The depths of 180 cm and  270 cm in  the

                              81

-------
ultisols  and  entisols represent approximate root zone  depths
in   these  soil  types.   The  900  cm  depth  represents   an
intermediate  depth to ground water in the ridge area.  Greater
depths  can  be much higher but occur with far less  frequency.
Areas  3,  4,  5,  and 6 have a  shallow,  relatively  constant
unsaturated zone, controlled largely by drainage practices.

These  combinations resulted in twenty scenarios.  In addition,
a  no-irrigation  simulation  was  made in  each  spodosol  and
alfisol  area.  This was done because, for the low volume spray
case,  part  of the treated band receives only rainfall.   Thus
the  pesticide  load  leached  from the grove  is  the  areally
weighted  average of the load from the wetted area  (irrigation
plus  rainfall) and the nonirrigated area (rainfall only).  For
ease  of  referring to these scenarios later in the text,  they
have  been  assigned the codes shown in Table 2.7.   The  first
letter  indicates  the soil order; the next letter, the  method
of  irrigation;  the  following number the depth  of  the  soil
core;   and,  the  last  letter,  the  low  or  high   rainfall
condition.
2.1.7  Selection of PRZM Hydrology and Hydraulic Input
       Parameters
For  each  unsaturated zone scenario  (1-6) input data sets  had
to  be  created  for PRZM.  The following  sections  cover  the
specification  of  all  the required hydrologic  and  hydraulic
input   parameters.   Most  were  selected  based  on  guidance
provided  in  the  PRZM  user's guide  (Carsel  et  al.,   1984).
Exceptions are listed and discussed.
                              82

-------
        TABLE 2.7  UNSATURATED ZONE MODELING SCENARIOS
 Scenario
Identifier
                               Description
Soil
Irrigation
Core Depth(m)
EN2.7
EN9.0
EO2.7
EO9.0
EL2.7
EL9.0
UNI. 8
UN9.0
U01.8
U09.0
UL1.8
UL9.0
SN1 . 2L*
S01.2L
SL1.2L
SN1 . 2H
S01.2H
SL1.2H
AN1.2L
AO1 . 2L
AL1 . 2L
AN1 . 2H
AO1 . 2H
AL1 . 2H
Entisols
Entisols
Entisols
Entisols
Entisols
Entisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
Alfisols
Alfisols
Alfisols
Alfisols
Alfisols
Alfisols
None
None
Overhead
Overhead
Low volume spray
Low volume spray
None
None
Overhead
Overhead
Low volume spray
Low volume spray
None
Overhead
Low volume spray
None
Overhead
Low volume spray
None
Overhead
Low volume spray
None
Overhead
Low volume spray
2.7
9.0
2.7
9.0
2.7
9.0
1.8
9.0
1.8
9.0
1.8
9.0
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
*L or H in the last position refers to low or high rainfall
 areas.
                              83

-------
2.1.7.1  Simulation Period—

The  six locations appearing on the map of Figure 2.14 show the
stations  for which meteorological data was initially obtained.
The  stations  were primarily chosen for their coverage of  the
citrus  growing area and the availability of precipitation, pan
evaporation  and  temperature data.  These data sets  were  put
into the proper format for PRZM.

Twenty-five  years  of precipitation and pan  evaporation  data
were  available  at most of these six locations.  However,  for
some  stations  pan evaporation data was  missing.   Therefore,
scenarios   were  set  up  to  run  for  only  14  years.   The
meteorologic  data  set chosen for entisols, ultisols, and  low
rainfall  spodosols and alfisols was the Lake Alfred Experiment
Station.   The  vhigh rainfall spodosol and alfisol  cases  were
run with Belle Glade meteorological data.

2.1.7.2  Hydrologic Parameters—

The  evaporation  pan coefficient was set at 0.78.   The  ANETD
parameter  is meaningless here since a crop is always  present.
Percentage  of  daylight hours (DT(12) were input  because pan
evaporation  data  were  sometimes missing for  short  periods.
They  were chosen based on a latitude of 28  N.  The snow  melt
coefficient  was set to zero.  The runoff curve number for  AMC
II  was  set at a very low value of 20 so that no runoff  would
occur.

2.1.7.3  Crop Parameters—

The  crop  was  assumed  to be  mature  throughout  the  entire
simulation  period.  Trees were assumed to have an interception
potential  of 0.15 cm.  This was based on information found  in
Viessman  et al., 1977.  The root  zone depth of the citrus tree

                              84

-------
     ALABAMA
                                          Jacksonville
           Gulf of Mexico
                                                 Atlantic Ocean
                                         Lake Alfred
                                       Vero Beach
                                      •Clewiston
                                         Belle Glade
Figure 2. 14   Selected meteorologic  stations  for
               jinsaturated gone  modeling.
                        85

-------
was  set at 240 cm (8 ft) and 150 cm (5 ft) in the entisols and
ultisols  respectively,  and at 60 cm (2 ft) in  the  spodosols
and  alfisols (Tucker, 1978).  The cover parameter (COVMAX) for
entisols  and  ultisols was estimated to be 75%.  This  is  the
approximate   area  of  a  7.6  m  (25  foot)  diameter  circle
inscribed  in  a 7.6 by 7.6 m square, thus assuming  that  tree
canopies  just  touch  on  a 7.6 by 7.6  m  spacing.   For  the
spodosols  and  alfisols,  on which tree canopies  are  smaller
(4.6  m diameter) and plantings are closer (usually 7.6 by  4.6
m) the cover parameter was set at 50%.

2.1.7.4  Soil Hydraulic Parameters—

Selection  of soils parameters was primarily a function of soil
group.  Therefore they are discussed by group below:

2.1.7.4.1   Entisols  and Ultisols—Entisols and ultisols  were
divided  into  those having thick and thin  unsaturated  zones.
For  the former a core depth of 900 cm  (30 ft) was used and for
the  latter  270 cm  (9 ft) or 180 cm (6 ft) was  used.   Thirty
compartments  were  used  in  each  core.   The  original  PRZM
hydraulics  module (HYDRl) was selected for these soils.   This
method  requires  input  values of field capacity  and  wilting
point  water contents and assumes that water in any compartment
above field capacity drains to that level within one day.

For  entisols, the profile was divided  into five horizons based
on  the  soil  characterization analysis performed as  part  of
this  study.  For ultisols, seven horizons were chosen.   Table
2.8  shows  the representative values of  8 fc and  9 ^ used  for
each horizon.

2.1.7.4.2   Spodosols  and Alfisols—For these two soil  types,
which  typically occur in the flatwoods, only one core depth of
120  cm  (4  ft)  was  used.  This  is  because  the  depth  to

                              86

-------
 TABLE 2.8  HYDRAULIC CHARACTERISTIC DATA FOR FLORIDA SOILS BY
            HORIZON
Soil
Group
Entisols
Entisols
Entisols
Entisols
Entisols
Horizon
1
2
3
4
5
Depth
(cm)
0-10
10-20
20-40
40-100
100+
Bfc
5.6
5.4
4.2
3.5
3.0
9wp
2.2
2.1
2.0
1.5
1.2
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols

Alfisols
Alfisols
Alfisols
Alfisols
Alfisols

Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
1
2
3
4
5
6
7

1
2
3
4
5

1
2
3
4
5
6
0-20
20-40
40-100
100-160
160-200
200-240
240+
0-20
20-40
40-80
80-120
120-240
0-10
10-40
40-60
60-100
100-200
200+
8.2
7.0
5.5
8.0
20.0
22.5
5.0
10.0
6.5
9.5
20.0
25.0
14.0
9.0
6.0
12.0
21.0
6.0
3.0
2.5
1.8
4.0
13.0
16.0
1.8
3.5
6.0
5.5
9.5
12.5
4.2
3.2
2.2
4.0
7.8
2.0
                              87

-------
saturation  in these areas is not controlled by relief, but  by
the  control  of  the  water table  through  drainage.   Thirty
compartments were again used to simulate each core.

In   the   spodosols   and  alfisols,  the  presence   of   low
permeability  soil  layers  or the presence of the  high  water
table  may  cause  both temporary saturated conditions  in  the
unsaturated  zone  and lateral drainage into adjacent  drainage
furrows.   For  this reason the HYDR2 option was used for  soil
hydraulics  with  lateral  drainage to  more  closely  simulate
recharge   to   the  saturated  zone.   In  addition   to   the
specification   of  field  capacity  and  wilting  point  water
contents  for  each horizon (Table 2.8), this  method  requires
that  time  constants  for  vertical and  lateral  drainage  be
input.

Unfortunately,   these  constants  are  not  easily  estimated.
There  are  two ways that the movement of water vertically  can
be  limited; either by the presence of a low permeability layer
or  the  presence of a saturated zone.  If the permeability  is
limiting  the  time  constant  for  vertical  drainage  can  be
estimated  from  the saturated hydraulic conductivity.  If  the
presence  of  a  saturated  zone is the  limiting  factor  then
downward  vertical  movement is limited by the rate of fall  of
the water table.

Inspection  of  the saturated hydraulic conductivities  in  the
low  permeability layers in spodosols and alfisols (see  Figure
2.7)  indicates that, while hydraulic conductivity is  far lower
in  these horizons than in the remainder of the profile, lowest
values  still  are  on the order of 1 cm/hr (on  the   average).
Thus,  it  appears that while some profiles may exist  in  which
the  vertical  movement is restricted by permeability, for  the
most  part,  vertical drainage is limited by the presence of  a
saturated  zone.   Therefore the rate of movement of water  and
                              88

-------
pesticide  downward is controlled by localized manipulation  of
the  water  table.   This  fact makes  the  time  constant  for
vertical  drainage  even harder to assess for a  general  case.
Similarly,  the  rate of lateral drainage is  almost  certainly
controlled  by the manipulation of water depths in the furrows,
rather than soil permeabilities.

The  only  data located which could be used to calculate  these
constants  was that of Calvert et al. (1967).  Figure 2.15 from
that  report  shows  the fall of a water table  in  two  bedded
citrus  groves.   On the right is an Immokalee soil  (spodosol)
and  on  the left is a Felda soil (alfisol).   Specific  yields
(saturation  minus  field  capacity water  contents)  for  each
horizon  in  each  soil were determined from profiles  for  the
respective  soil types found in Calhoun et al. (1974), Carlisle
et  al. (1978) and Carlisle et al. (1981).  Water draining from
0-22  hrs.  in  the Immokalee soil and from 0-25  hrs.  in  the
Felda  soil was assumed to all move laterally.  Time  constants
for  lateral drainage were determined by dividing the depth  of
water  drained by the time of drainage and the depth of a  PRZM
soil  layer  (in this case 4 cm).  Time constants  for  lateral
drainage  were thus determined to be 0.48 /day for the  alfisol
and 0.21 for the spodosol.

Similarly,  vertical  time  constants were  estimated  for  the
drainage  of  the remainder of the water in the  profile  which
was  assumed  to  all move vertically.   These  constants  were
found  to  be 0.065 /day for the alfisol and 0.32 /day for  the
spodosol.

In  the  PRZM  input data set, horizons above a 60  cm   (2  ft)
datum  were assigned lateral drainage parameters equal to those
calculated;  horizons  deeper than 60 cm were assigned  lateral
drainage  parameters of zero.  The vertical drainage parameters
calculated  above  were assigned to the deepest  PRZM  horizon;

                              89

-------
  (0
  w
  g
w
Q
 o-

 6-

12-

18-

24-

30

36-

42-
          FELDA SOIL
                  Ohr
                0+25 hrs
           	


        I 0+241 hrsJV--^^  ^-~J

           IMPERVIOUS CLAY
                LAYER
           I
           0
                                      IMMOKALEE
                                         SOIL
                                 0 hr
                               0+2 2_ hrs_

                               0+115 hrs
               I      I
               5     10    15     10     5

              DISTANCE IN FEET FROM WATER
                                           I
                                           0
Figure 2.15 Fall-of free water level during one  furrow
            irrigation drying cycle in single bedded
            'Ruby.Red1 grapefruit  groves planted on
            Felda and Immokalee  soil types in the
            Indian River area.   After Calvert et al.,
            1967.

             (1  inch =2.54 cm)
                        90

-------
overlying  horizons  were  assigned high  values  (10./day)  to
essentially provide unrestricted drainage.

2.1.7.5  Irrigation Parameters—

The  only  irrigation  parameters required  by  the  algorithms
developed  for  PRZM  (see Appendix A) are the  rate  of  water
application  and  the percent available water content at  which
events  are  triggered.   The water application rates  are  not
extremely  important.  Since PRZM operates on a daily timestep,
the  durations  of  the  events  (i.e.,  the  amount  of  water
necessary  to  bring the root zone profile from present  status
to  field  capacity divided by the input application rate)  are
rounded  to  the nearest whole day.  The application rates  are
adjusted  accordingly.   Even so, the rates shown in Table  2.5
were used for the irrigation system being simulated.

The  available  water  level at which  irrigation  events  were
triggered  were  set  initially, at 50% but  were  adjusted  to
achieve  an annual evapotranspiration depth of about 122 cm (48
in.)  for  the  overhead irrigation method.  Final  values  for
this parameter was 45% for all soil orders.
2.2  SATURATED ZONE
The  goal of the saturated zone modeling study was to  estimate
probable  aldicarb concentrations in drinking water wells.   In
general   terms,   there   are  three   environmental   factors
influencing   the   potential  movement  of  aldicarb  in   the
saturated  zone:   1)  the  geometry or  configuration  of  the
aquifer  system, 2) the physical properties of the aquifers and
confining  layers,  and 3) the influence of the drinking water
                             91

-------
well.   Table 2.9 summarizes the aspects of these factors  that
are most important.
2.2.1  Aquifers and Aquifer Geometries
The  occurrence  and  movement of ground water  in  the  citrus
growing  area  of Florida are closely related to  its  geology.
The   central  Floridan  peninsula  is  comprised  of  a  thick
sequence   of  hydrologically  connected  limestone  formations
which  make  up  the principal artesian aquifer,  the  Floridan
Aquifer.   This  is  overlain by younger  alluvial  and  marine
deposits  which  contain the unconfined surficial  aquifer  and
the  intermediate confined aquifers.  All three are present  in
the  citrus  growing  areas and supply drinking  water.   Table
2.10   summarizes   the  hydrologic  and  geologic   properties
discussed below for these aquifers.

2.2.1.1  Floridan Aquifer—

The  Floridan Aquifer is one of the largest and most productive
aquifers  in  the  world.   It underlies the  entire  state  of
Florida and parts of Georgia, South Carolina, and Alabama.

Parker  et  al. (1955) originally defined the Floridan  Aquifer
as:

    "...parts  or  all of the middle Eocene  (Avon  Park  and
    Lake  city  Limestone), upper Eocene (Ocala  Limestone),
    Oligocene   (Suwannee  Limestone),  and  Miocene   (Tampa
    Limestone   and   permeable  parts  of   the   Hawthorne
    Formation  that are in hydrologic contact with the  rest
    of the aquifer)."
                             92

-------
TABLE 2.9  PHYSICAL FACTORS INFLUENCING PESTICIDE MOVEMENT IN
           THE SATURATED ZONE
             1.  Aquifer Geometries


                   a.  Floridan Aquifer

                   b.  Intermediate aquifer

                   c.  Surficial aquifer

                   d.  Physical relationship between
                       aquifers


             2.  Hydrogeologic Properties

                   a.  Hydraulic conductivity

                   b.  Hydraulic gradient

                   c.  Porosity

                   d.  Thickness of aquifer

                   e.  Recharge to saturated zone


             3.  Drinking Well Specification

                   a.  Distance between pesticide source
                       area and well

                   b.  Well depth

                   c.  Well rate
                              93

-------
            TABLE    2.10
             GENERALIZED STRATIGRAPHIC  UNITS AND ASSOCIATED HYDROGEOLOGIC
             PROPERTIES  (Compiled  from  Stringfield,  1966;  Klein  et al.,
             1964;  Knochenmus  and  Hughes,  1976;  Grain et  al.,  1975;  Tibbals,
             1981;  Healy,  1982;  Merritt et  al.,  1983).
GEOLOGIC ACE
SERIES
STRATAGRAPHIC
UNIT
RANGE LITHOLOGY
OF
THICKNESS (f t)
HATER-BEARINO
• -PROPERTIES
HYDROLOGIC
UNIT
     Holocene
     Pleistocene
vo
     Pliocene
     Miocene
                   Unnamed alluvial,  lake
                   and windblown deposit*
                       0-70       Alluvium, freshwater marl
                                  peats and muds in stream •
                                  and lake bottoms.  Also
                                  some dunes and other wind-
                                  blown sand
Relatively impermeable
                  Pamlieo formation
                  and marine and estua-
                  rine terrace deposits
                       0-100      Mostly marine quarts  sand,
                                  unconsolidated, and generally
                                  well graded.  Also, some
                                  fluviatile and lacustrine
                                  sand, clay, marl, and peat
                                  deposits
Generally  low permeabilities
Fort Thompson Formation  50-150
(contemporaneous with
Anastasia)
                                                    Alternating marine shell beds
                                                    and fresh-water marl
Generally of low permeabilitj.es
except locally where it is
solution riddled
                  Anastasia Formation
                                        50-150
                                                    Sand, marl,  and shell beds
                                                             Moderate to high permeability
                                                                                                                      1
                  Citronelle Formation
                                         0-200      Marginal marine fine to
                                                    course grained quarts sand
                                                    containing kaolin!te matrix,
                                                    variegated red and orange
                                                    quartzite pebbles, cross-
                                                    bedded
                                                             Yields  large quantities of
                                                             good quality water
                  Caloosahatchee Marl
                                        0-60
                                                    Shell, sand, silt and-marl
                                                             Shell beds yield water that in
                                                             some areas is highly mineralised
                  Alachua and Bone
                  Valley Formations
                      0-100        Nonmarine interbedded depo-
                                  sits of clay, sand, and sandy
                                  clayi much of unit is phos-
                                  phatic, base characterised by
                                  rubble of phosphate rock and
                                  silicified limestone residuum
                                  in a gray and green phosphatio
                     	clay matrix    	
Yields small to moderate
quantities of water to shallow
wells

H
£
                  Tamiami Formation
                                        0-150
                                                    Sand, marl,
                                                    limestone
                                            shell beds, and
Upper part of formation is very
permeable.  Lower part, with the
Hawthorn,  confines water in the
underlying principal artesian
aquifer    	        	
                                                                                      (Continued)
                                                                                                   5

-------
      TABLE   2.10  (continued)
vo
SERIES
Miocene
Oligocene
Upper
Eocene a
I
o
a
r*
a
8
Middle Eocene
BTRATIGRAPHIC
UNIT
Hawthorn Formation
Tampa Limestone
Suwanee Limestone
Crystal River
Formation
Hillleton and
Inglia Formations
Avon Park
Limestone
Lake City Limestone
RANGE LITHOLOGY
OF
THICKNESS (ft)
0-550 Marine interbedded mixture
of sand, clay, and lime-
stone i fine to course
grained phosphatic sand, .
dark green to cream mont-
morillonitlc clay, phos-
phorite pebbles, lenses of
sandy dolooitic phosphatic
limestone
0-250 Fossilferous limestone with
sands, silts and clay. Some
parts of the limestone is
silicified
0-450 Marine limestone, cream to
white, soft, hard where
silicified, porous
25-300 Gray to cream colored porous
massive limestone i often
consists of tests of foramin-
ifers, cherty, in places
25-125 Tan to cream colored granular,
porous limestone, highly
fossiliferous, lower part at
places is dolomite
400-1000 Marine limestone, light gray
to dark brown, soft to hard,
fossiliferous, dolomitic,
carbonaceous, contains gypsum
and chert
200-1000 Alternating beds of dark brown
dolomite and chalky limestone
• minor amounts of gypsum and
anhydrite
HATER-BEARING
PROPERTIES
Yields small to moderate
quantities of artesian and
nonartesian water. Major
part of Hawthorn forms the
confining layer for the
underlying artesian water,
but lower part forms the
upper part of the principal
artesian aquifer
Yields large quantities of
water in west-central Florida,
but is generally lower in
permeability in other parts of
the state except locally in
discrete solution zones
Yields moderate amounts of water
but generally less than under-
lying Eocene formations.
Contains solution cavities in
recharge area
One of most productive forma-
tions of the Floridan Aquifer.
Generally yields good quantity
of water .
Hater yields generally less than
the Crystal River Formation.
Contains many solution cavities
in recharge areas
Principal source of water in
areas where overlying limestone
is thin or absent. Yields water
from porous xones highly
mineral! led in many areas
Hater yields varies, important
source of artesian water in
some areas
HYDROLOGIC
UNIT
[< Intermediate Aquifers 	
a \
I \
L
SM
5&
s|
r
•4
t
4
t
1
\
1
4
      Lower Eocene   Oldsmar Limestone
                                          400-600      Marine Limestone, light brown  Contains salt water
                                                      interbedded with brown crystal-
                                                      line dolomite,  fossiliferous,
                                                      contains chert  and gypaum

-------
It  is  confined above by the low permeability clays and  marls
of  the  Hawthorn  and  Tamiami  Formations.   The  aquifer  is
generally  fossiliferous and porous, ranging in thickness  from
100  m (300 ft) to 600 m (2000 ft).  Parts are highly fractured
and  riddled  with solution cavities.   In some areas  the  Avon
Park  Limestone  is  of  lower  permeability  and  divides  the
aquifer into upper and lower permeable zones  (Bush, 1982).

In  the  northwestern  portion of the study  area,  in  Citrus,
Hernando,  Sumter  and  Pasco counties, there is  a  structural
high  where  the limestone formations in the upper  portion  of
the  Floridan Aquifer outcrop at the land surface (Stringfield,
1966).   In  this area the Floridan Aquifer is  unconfined  and
there  is  a  direct  connection between the  surface  and  the
aquifer.  This means water and contaminants can move much  more
quickly into the aquifer than where a confining layer exists.

The  limestone  strata gently dip and thicken to the south  and
southeast  of the structural high.  In these areas the Floridan
Aquifer  is  overlain by the marine and montmorillonitic  clays
in  the Hawthorn Formation and marls in the Tamiami Formations.
VThen  the  Floridan Aquifer is confined, water will rise  above
the  top of the aquifer in a well.  As long as the water  level
is  below the level of the water table there will be  potential
for  downward movement of water.  If, however, the water  levels
in  the Floridan are higher than the water levels in the  upper
aquifers  there will be no driving  force downward and therefore
no  chance  for  the migration of aldicarb  into  the  Floridan
Aquifer.  These areas are shown on  Figure 2.16.

In  some  areas,  especially the central  highlands,  sinkholes
have   breached   the  confining  layer,  creating   a    direct
hydrologic  link  between the surface and Floridan Aquifer.    A
sinkhole  occurs when a solution cavity becomes so big the roof
can  no  longer support the overburden.  The  collapse   feature

                             96

-------
         AIABAMft
                                                    Jacksonville
                                                        Atlantic Ocean
Figure 2.16  Principal Areas where the piezometric surface of the
             Floridan Aquifer rises above the water table.
             Source:  Healy, 1975.
                                97

-------
sometimes  fills up with water, creating a lake.   Occasionally
the  low  permeable  lake  bottom  sediments  plug  the  direct
conduit  to  the ground water.  Often, however,  the  sinkholes
remain  open, as a direct link to bring water and  contaminants
into the aquifers.

Not  all  parts of the Floridan Aquifer contain potable  water.
In  much of the coastal area the Floridan is not potable due to
salt-water  contamination.   In  these areas (shown  in  Figure
2.17)  aldicarb contamination to the Floridan does not need  to
be  considered, since the drinking water is supplied from other
aquifers.

In  summary,  for the Floridan Aquifer, the highest risk  areas
for  pesticide contamination are in the northwestern area where
it  is unconfined and in the ridge area where sink holes create
direct  links  with  the surface.  Where the  Floridan  Aquifer
does  not  meet drinking water standards and where there is  no
potential  for downward leakage, the Floridan Aquifer does  not
have  to  be considered.  Where the Floridan is  confined,  the
risk  depends  on  how quickly water and  pesticide  will  move
downward   through  the  overlying  aquifer  or  aquifers   and
confining  layers.   The  speed of contaminant  transport  will
depend on the hydrogeologic parameters.

2.2.1.2  Intermediate Aquifers—

The  intermediate  confined aquifers are in  general  localized
and  not very laterally extensive.  They occur  in the permeable
shell  beds and limestone lenses that  vary in thickness from   3
to  60+  m   (10  to  200+  ft)  in  the  Tamiami  and  Hawthorn
Formations  of  Miocene age  (Healy, 1982).  These aquifers  are
usually  confined  above and below by  the impermeable parts  of
the  same geologic formations.  In Polk County  an  intermediate
confined  aquifer  occurs in the younger Pliocene  deposits  of

                              98

-------
     ALABAMA
                                                 Jacksonville
                                                      Atlantic Ocean
Figure  2.17  Areas of  the Floridan  Aquifer which  do not meet
              drinking  water standards.
              Source:   Klein, 1975.
                             99

-------
pebbly  phosphatic sands in the Bone Valley Formation (Stewart,
1966).   Some  authors consider the upper limestone  member  of
the  Hawthorn Formations an intermediate aquifer distinct  from
the  Floridan.  Over the southern two-thirds of Polk County and
all  of DeSoto and Hardee counties, this permeable limestone is
hydrologically  separate  from  the Floridan with  only  a  few
localized connections (Steward, 1966 and Wilson, 1977).

The   intermediate  aquifers  exist  most  extensively  in  the
south-central  part of the study area, from Polk County  south.
They  are tapped v/hen the Floridan Aquifer is not suitable for
drinking  water  and the unconfined surficial aquifer does  not
yield large enough quantities of water.

2.2.1.3  Surficial Aquifer—

An  unconfined,  surficial aquifer occurs in most areas of  the
state  in the uppermost, unconsolidated deposits that range  in
age  from  Miocene to Holocene.  It is extremely variable  both
in  composition  and  thickness.   In most  areas  the  aquifer
averages  6 to 9 m (20 to 30 ft) in thickness, but is as little
as  .3  m (1 ft) in parts of Brevard, Highlands,  Manatee,  and
Palm  Beach  counties,  and as much as 120 m  (400  ft)  in  St.
Lucie  and  Indian River counties.  It is composed  chiefly  of
sands,  shells,  limestone, clay and marl.  The geologic  units
that  contain  the unconfined aquifers depend on the  structure
and stratigraphy of a particular area.

In   the   central   ridge  area  the  surficial   aquifer   is
discontinuous   and   extremely  variable  in  thickness. .   In
southeastern  Polk County, for example, the unconfined  aquifer
varies  from  7.5  to  75 meters (25 to 250   ft)  in  thickness
(Wolansky et al., 1979).

In  the  coastal  citrus  area the surficial  aquifer   is  much

                             100

-------
thicker  and  more  laterally  continuous.  It  occurs  in  the
extremely   variable  Pamlico,  Fort  Thompson,  and  Anastasia
Formations   of  Pleistocene  age  (Healy,  1982).   They   are
composed  chiefly  of marine sands and shell beds,  interbedded
with clay and marl deposits.

In  the  ridge  area the surficial  aquifer  supplies  drinking
water  for small domestic users.  The surficial aquifer in  the
coastal   citrus  areas  provides  the  principal  supply   for
drinking  water  for both municipalities and for  small  single
farms.

2.2.1.4  Aquifer Geometries—

For  the purposes of modeling, this very complicated system  of
aquifers  must  be simplified. The first simplification  is  to
only  consider the Floridan Aquifer where it meets the drinking
water  standards  and  where there is  potential  for  downward
leakage   of  water  (and  aldicarb)  to  the  Floridan  (shown
previously in Figures 2.16 and 2.17).

Figure  2.18  shows  the  delineation of  the  typical  aquifer
geometries  considered  in this study.  Areas  1, 2, and  3  are
areas  that  include  the Floridan Aquifer system.  Area  1  is
where  the Floridan is unconfined and considered alone.  Area 2
is  where a surficial unconfined aquifer overlies the  Floridan
with  a  confining layer in between.  Together they comprise  a
leaky  two-aquifer  system.   In area 3, a three-aquifer  system
exists,  consisting of the unconfined surficial aquifer, and  a
confined intermediate aquifer overlying the Floridan Aquifer.

In  areas  4 and 5 the Floridan is not included because of  its
poor  water  quality or because its potentiometric  surface  is
above  the  water table.  Area 4 is where only  the  unconfined
surfical  aquifer  is  considered.   Area 5  includes  both  an

                              101

-------
         ALABAMA
                                           Jacksonville
            Gulf of Mexico
                                          Atlantic Ocean
   1 -


   2 -



   3 -




   4 -


   5 -
     Explanation

Floridan Aquifer Alone
 - Unconfined

Leaky Two-Aquifer System
 - Surficial Aquifer
 - Floridan Aquifer

Leaky Three-Aquifer System
 - Surficial Aquifer
 - Intermediate Aquifer
 -  .Floridan Aquifer
Surficial Aquifer Alone
 - Unconfined
Two-Aquifer System
 - Surficial Aquifer
 - Intermediate Aquifer
Figure 2.18  Principal aquifer  geometries in the
             citrus growing area.
                           102

-------
unconfined   surficial  aquifer  and  a  confined  intermediate
aquifer.

Table  2.11 shows the approximate citrus acreage in each  area.
The  largest  amount  of citrus is grown in Area 4,  where  the
surficial  aquifer  is considered alone.  This is  followed  by
Areas  2,  3,  and 5 where a multi-aquifer system  is  modeled.
Area  1  has the smallest.amount of citrus grown  (only  3.0%),
yet   it   is  important  to  consider  because  the  risk   of
contamination   is  high.   Remaining  areas  have  substantial
quantities  of  citrus and thus should not be  eliminated  from
consideration.

These  five  geometries are grouped into three general  aquifer
configurations  for  modeling:   1) the Floridan, as  a  single
unconfined  aquifer (Area 1), 2) the unconfined surficial (Area
4),  and  3) a general two-aquifer system (Areas 2, 3, and  5),
be  it  the  surficial  aquifer overlying the  Floridan  or  an
intermediate confined aquifer.

The   three-aquifer   system  would  only  be   considered   if
significant   contamination  is  found  in  the  lower  of  the
two-aquifer system.
2.2.2  Hydrogeologic Properties
The  hydrogeologic properties in each of the aquifers determine
the  ground-water flow regime.  They describe the properties of
the  porous medium that collectively affect the velocity of the
water  and transport of aldicarb in the aquifers and  confining
layers.   The  five principle hydrogeologic properties used  in
the  saturated  zone  modeling  are each  defined  and  briefly
described.

                             103

-------
TABLE 2.11  APPROXIMATE CITRUS ACREAGE ASSOCIATED WITH EACH
            AQUIFER GEOMETRY
County
(in ranked
order)
Polk
Lake
St. Lucie
Indian River
Orange
Hardee
Martin
Hillsborough
DeSoto
Pa sco
Hendry
Osceola
Brevard
Manatee
Palm Beach
Marion
Volusia
keechobee
Lee
Hernando
Seminole
Charlotte
Glades
Pinellas
Surater
Citrus
Total
%
Area
123
31,000 63,000
87,000

10,000
28,500
42,000

14,000 17,500
16,500
13,000 17,000

12,000

8,000

2,000 2,000
2,500
2,000

5,000 700
600


2,000
600 1,000
1,000
21,600 212,300 147,000
3.0 29.5 20.4

4

17,000
64,000
50,000
14,000

32,000
3,500
13,000


2,500
15,000
5,000
7,500

4,500

5,000

5,000
4,500


1,300

243,800
33.8

5
31,000

7,000



6,500

3,500

30,000
2,500


5,000


4,500
1,000


1,000
4,000



96,000
13.3
                           104

-------
    1.    Hydraulic conductivity (L/T)  is  a measure of the  media
         properties  that  determine the  ease with which  water
         is   transmitted   through  a porous  medium.     High
         conductivities  mean  it  is  easily  transmitted,   low
         conductivities mean water moves  slowly.

    2.    The hydraulic gradient (L/L)  is  the vertical change in
         water  level  over a given horizontal  distance.    The
         gradient  acts  as the driving force to flow.   For  a
         given  porous  medium  the flow  will  be  greater  the
         larger (steeper) the hydraulc gradient.

                    3  3
    3.    Porosity (L/L ) is the percent  void space in a repre-
         sentative  volume of the porous  medium.  The  porosity
         affects  the  ground-water  velocity.   High  porosity
         material  has  lower ground-water velocities than  low
         porosity  material  for  the same  average  flux  rate
         (specific discharge).

    4.    Saturated thickness (L) of an aquifer is the thickness
         in  which  all the pores are filled with water.   When
         the  water  supply well is in the lower  aquifer,  the
         saturated  thickness  of  the unconfined  aquifer  and
         confining  layer  will influence the  vertical  travel
         time of the pesticide.

    5.    Recharge rate (L/T) is the average annual rate at
         which   water   is  replenished  to  the  aquifer   by
         percolation from the unsaturated zone.

Each  of  these parameters takes on a large range of values  in
the  aquifer  systems of central Florida.  Two combinations  of
these  property values were used for each of the three  aquifer
configurations  in the saturated zone modeling.  A "worst case"

                             105

-------
was  modeled  using  the extreme values  of  the  hydrogeologic
properties   that   collectively  move  aldicarb  through   the
saturated   zone  most  quickly.   The  other  combination   of
properties  will use intermediate values that best characterize
each of the three aquifer configurations.
2.2.3  Influence of Drinking Water Well
The  drinking water well influences the movement of the  ground
water  and  therefore  the  transport of  aldicarb.   The  well
position  relative  to the pesticide source area, the depth  of
the  well  and  the rate at which it pumps  all  influence  the
ground-water flow regime.

2.2.3.1  Well Distance from Aldicarb Application Site—

The  distance of a drinking water supply well from the area  of
pesticide   application  is  crucial  in  the  evaluation   and
assessment  of  potential  human exposure to  aldicarb  through
drinking  water.  The greater the distance necessary to  travel
from  the source area to the well through the ground water, the
smaller  the chance of contamination.  The distance of the well
from  the  source is an important management consideration  for
EPA.   The  current EPA standards prohibit the use of  aldicarb
within  91  m (300 ft) of a water supply well.   This  distance
was  evaluated  in all the scenarios.  The EPA  is  considering
300  m (1000 ft) as an alternative distance in their evaluation
of  the aldicarb regulations.  This distance was also simulated
in all the scenarios.

2.2.3.2  Well Depth—

In   the   Florida  citrus  growing  area,  water  wells   vary

                              106

-------
considerably   in   depth  depending  on  the   local   aquifer
characteristics  and  the water needs.  Deep wells  with  large
capacities   are  usually  drilled  to  supply  municipalities.
Shallow  wells  have  much smaller capacities  and  supply  the
small  domestic  user.  The depth that the water  is  withdrawn
from  a  well influences the localized aquifer flow system.   A
shallow   well  affects  the  near  surface  flow  system   and
therefore   would   influence  the  movement  of   a   leaching
contaminant  sooner than a deep well with an equivalent pumping
rate.

2.2.3.3  Well Pumping Rates—

The  well rate determines the magnitude of the influence on the
regional   flow  system.   A  low  pumping  rate  has  a  small
localized  effect.   High  pumping rates can  change  the  flow
system   on  a  much  larger  scale,  greatly  increasing   the
ground-water  velocities  towards the well.  Two well rates,  a
high  and  a  low  rate,  were  simulated  in  each  hydrologic
setting.
2.2.4  Delineation of Saturated Zone Scenarios
The  factors  influencing  aldicarb transport are  reviewed  in
Table  2.12  with  the  cases that  were  considered  for  each
factor.   All the combinations of these factors resulted in  48
modeling  scenarios.   This  number  doubled  to  96  potential
scenarios  because  of the consideration of two chemical  decay
rates.   Two  decay rates were used; 1) the best estimate of  a
representative rate, and 2) no decay.

The  following notation convention is used for identifying each
scenario.

                             107

-------
   TABLE 2.12  CASES CONSIDERED FOR SATURATED ZONE MODELING
Influencing Factors
Cases Considered for Each Factor
Aquifer geometry
1) Unconfined Floridan Aquifer
2) Unconfined surficial aquifer
3) Leaky two-aquifer system
Aquifer properties
1) Worst case
2) Intermediate case
Well distance
1) 300 ft
2) 1000 ft case
Well depth
1) Shallow
2) Deep
Well rate
1) High rate
2) Low rate
                             108

-------
    

                     Aquifer Geometries
                        F = Floridan Unconfined
                        S = Surficial Unconfined
                        2 = Two Aquifer System

                     Aquifer Properties
                        W = Worst Case
                        A = Average Case
                     Decay Rates
                        D (rate in days   )
                     Well Configurations
                        W (distance in feet, depth in feet,
                          rate in gpm)

FWD(0)W(300,50,2000),  for example, would identify the scenario
that  uses  the unconfined Floridan Aquifer with the worst  set
of  hydrogeologic properties, no decay rate, a well distance of
300  ft  from the aldicarb source, a well depth of 50  ft,  and
pumping rate of 2000 gallons per minute  (gpm).
2.2.5  CFEST Model Parameters
2.2.5.1  Hydrogeologic Parameters—

The  hydrogeologic parameters necessary for modeling with CFEST
detex'mine   the  ground-water  flow  regime  that  collectively
affect  the velocity of the water and transport of aldicarb  in

                             109

-------
the  aquifers  and  confining  layers.  The  equation  for  the
average  ground-water  velocity shows the relationship  between
the hydrologic properties used by the model:
          V =                                            (2.1)
where     Y  = average ground-water velocity
          K  = hydraulic conductivity
          I  = hydraulic gradient and
          e  = porosity

Hydraulic  conductivity,  hydraulic gradient and  porosity  are
the   most   important  properties  affecting  the   speed   of
ground-water  flow.   The saturated thickness of  the  aquifers
and   confining  layers  and  the  annual  recharge  rate  also
influence  the  hydraulics  of  the system  as  well  as  being
necessary  modeling parameters.  Each of these parameters takes
on  a  large range of values in the aquifer systems of  central
Florida.   The  worst and average set of  hydrogeologic  values
used  for  modeling for each aquifer system are shown in  Table
2.13.   The  determination of numerical values for  these  five
hydrogeologic parameters are discussed below.

2.2.5.1.1	Hydraulic  Conductivity—Hydraulic  conductivities
can   be   measured  in  the  field  or  in  the  lab.    Field
measurements  are better for modeling studies because they  use
several  wells  and test larger aquifer-scale properties.   Lab
tests  on the other hand are done on small samples.   Therefore
the    properties    measured    are    extremely    localized.
Unfortunately,  field  testing of hydraulic  conductivities  is
more  expensive and not as common as lab measurements.  A third
method  to determine the hydraulic conductivities is to model a
particular   area.   When  only  limited  data  are  available,

                             110

-------
TABLE   2.13 AVERAGE  AND  WORST CASE HYDROGEOLOGIC PARAMETERS FOR THE THREE
            AQUIFER  CONFIGURATIONS

Hydraulic Conductivity
Saturated Thickness
W Hydraulic Gradient
K Porosity
W
Recharge Rate
Floridan Aquifer
Unconfined
1000 ft/d
600 ft
2 ft/ 1 mile
.30
.20-. 35
13 in/yr
Surficial Aquifer
Unconfined
30 ft/d
75 ft
2 ft/1 mile
.20
.15-. 25
12 in/yr
TV
Surficial
23 ft/d
75 ft.
5 ft/1 mile
.20
.20-. 30
10 .in/yr
ro-Aquifer System
Confining Layer
lxlO-*ft/d
40 ft
AH = 35 ft .
.10
.05-. 10
10 in/yr
Floridan
200 ft/d
600 ft
4 ft/mile
.30
10 in/yr

Hydraulic Conductivity
Saturated Thickness
Hydraulic Gradient
Porosity
Recharge Rate
Floridan Aquifer
Unconf indd
5000 ft/d
600 ft
3 ft/1 .mile
.20
25 in/yr
Surficial Aquifer
Unoonfined
100 ft/d
75 ft
4 ft/1 mile
• 15
30 in/yr
TV
Surficial
40 ft/d
40 ft
25 ft/1 mile
.20 .
20 in/yr
/o-Aquifer System
Confining Layer
2. 5x10-* ft/d
10 ft
AH = 40 ft.
•05
20 in/yr
Floridan
1770 ft/d
100 ft
10 ft/1 mile
.20
20 in/yr
Conversions:
1 ft = .3048 m
1 in.  = 2.54 cm

-------
duplicating  the regional historical hydrologic conditions with
computer  simulation can often produce a good regional  picture
of  the hydraulic conductivites.  All three of these sources of
data   on   hydraulic  conductivity  were  used  to   determine
parameter values for this modeling study.

In  the  area where the Floridan is unconfined (Area 1),  Ryder
(1982)  has  done  a computer simulation  model  of  historical
conditions  in which he incorporated the existing field and lab
data  on  the  hydrologic  parameters.  This  provided  a  good
source   of  regionally  calibrated  aquifer  transmissivities.
From    these   and   the   saturated   thickness,    hydraulic
conductivities  were  calculated, ranging from several  hundred
                  — 4
ft/day  (7.06 x 10   m/s) to over 5000 ft/day (.02 m/s).  These
were  broken down into an average value of 1000 ft/day  (3.53  x
10    m/s)  and worst case value of 5000 ft/day (0.02  m/s)  as
shown in Table 2.12.

The  surficial  aquifer is unconfined and considered along  all
the  east  and  west coasts (Area 4).  Unfortunately  there  is
very  little  technical data on the surficial  aquifer.   Scott
(1977),  for  example,  reports hydraulic  conductivities  that
                                       -6             -4
range  from  1  to  130 ft/day (3.5 x 10  to 4.6 x  10     m/s).
These  values  were calculated from lab tests and a  few  field
tests  to set the limits.  Intermediate values were  calculated
using  lithologic   and physical properties. In  Martin  County,
Lichtler  (1960)  reports  field pump test results  from  which
hydraulic  conductivities  can be calculated which  range  from
about  20  to  150  ft/day  (7.1 x 10 ~5 to 5.3 x  10"   m/s)  and
averaging  40  ft/day  (1.4 x 10   m/s).  Values  reported  and
calculated  for  St.  Lucie County  (Bearden,  1972),  Charlotte
County  (Wolansky,  1978)  and Lee County (Wedderburn   et  al.,
1982) fall into this same  range.

A  two-aquifer system is modeled in the central portion of  the

                             112

-------
peninsula  (Areas 2, 3 and 5 in Figure 2.18).  For this  system
hydraulic  conductivity  is needed for the confining  layer  as
well  as  the  surficial unconfined aquifer  and  the  confined
aquifer.  Data for hydraulic conductivity in both the surficial
aquifer  and  confining  layer are  limited.   Lichtler  (1972)
reports  values ranging from 5 ft/day to 40 ft/day (1.8 x  10
to   1.4  x  10     m/s)  for  the  surficial  aquifer  in  the
east-central  area.  Two pump tests in the surficial aquifer in
the  Polk  County region showed hydraulic conductivities of  55
and  32  ft/day  (1.9 x 10~4and 1.1 x 10 ~4  m/s)  (Hutchinson,
1978).   Hydraulic conductivities estimated from  transmissivi-
ties  in  DeSoto  and  Hardee Counties (Wilson,  1977)  and  in
west-central  Florida  (Ryder,  1982) are in  the  same  range.
Weighted  averages  based  on the quality of the data  and  the
quantity  of  citrus  resulted in the average  and  worst  case
values shown in Table 2.13.

A   similar  process  was  used  to  determine  the   hydraulic
conductivity  of  the  confining  layer.   Wilson  and  Gerhert
(1982),  in  their  modeling  study  of  west-central  Florida,
specifically   calculated  vertical  hydraulic  conductivities.
Values  from their study were compared to values estimated from
leakances  in  the middle gulf area  (Cherry et al.,  1970)  and
Highland  County (Bishop, 1956).  The resulting worst case  and
average case values are shown in Table 2.13.

The  confined aquifer in the two-aquifer system could be either
an  intermediate aquifer or the Floridan Aquifer.  Because  the
intermediate  aquifers  are  usually  localized  and  not  very
laterally  extensive,  the properties of the  Floridan  Aquifer
are  used  to represent the confined aquifer.   Three  regional
modeling  studies  (Tibbals, 1982; Wilson and Gerhert, 1982; and
Ryder,  1982) were used to determine the average and worst case
hydraulic  conductivities.   These  values  compared  favorably
with  values  estimated from transmissivities reported  in  the

                             113

-------
local county reports sited above.

2.2.5.1.2   Hydraulic  Gradient — The hydraulic gradient is  the
vertical  change in hydraulic head (water level or  piezometric
surface)  over  a given horizontal distance.  In the  confining
layer  the  driving force is the difference in  hydraulic  head
between  the  upper  and lower aquifers.  The gradient  is  the
major  driving force for ground-water flow.  For a given porous
medium  the  flow  will  be greater the  larger  (steeper)  the
hydraulic  gradient.   The hydraulic gradient can  be  measured
directly  on  a map of the water table or piezometric  surface.
Figure  2.19  shows  the piezometric surface  of  the  Floridan
Aquifer  throughout  Florida.  A more detailed picture  of  the
hydraulic   gradient  where  the  Floridan  is  unconfined  was
obtained   from  a  regional  modeling  study  (Ryder,   1982).
Regional  studies  in the central peninsula where the  Floridan
is  confined (Hutchinson, 1978; Cherry et al . , 1970; and Wilson
and  Gerhert,  1982) helped to estimate the extreme  values  of
the hydraulic gradient.

The  same approach was used to determine the hydraulic gradient
for  the  surficial aquifer.  Healy (1982) has  compiled  water
level  maps for the surficial aquifer in the areas where it  is
the  principal  aquifer used for drinking water (equivalent  to
Area   4).   Bearden   (1972)  showed  water  level  maps   with
gradients  up to 7.58 x 10   (4 ft/mi.).  Most gradients ranged
from  2.84  x  10~4(1.5 ft/mi.) to 6.63 x  10 ~4  (3.5  ft/mi.)
(Land  et  al., 1973; Lichter, 1960; and Grain et  al . ,  1975).
Where  the surficial aquifer is part of the two-aquifer  system
in  the ridge area the gradients are much more variable due  to
      more  variable  topography.   Hutchinson,  1978,  reports
                                     — 4
hydraulic  gradients  from 4.73 x 10    (2.5 ft/mi.) to  4.73  x
10 ~^  (25 ft/mi.).  Most areas have gradients that average 9.47
x  10 ~4   (5 ft/mi.) (Cherry et al., 1970; Sinclair,   1974;  and
Wilson and Gerhert, 1982).

                              114

-------
                                              Jacksonville
                                                    Atlantic Ocean
        Datum is Mean  Sea Level
	20— Potentiometric Contour, in  feet
 Figure 2.19  Potentiometric  surface of the Floridan Aquifer
              After:  Healyr  1975.
                              115

-------
In  the confining layer the driving force is the difference  in
hydraulic  head  between the upper and lower  aquifer.   Wilson
and  Gerhert (1982) present generalized hydraulic head maps for
both  the  upper  and  lower aquifers.   From  these  maps  the
difference  in heads is estimated to vary between 0 and 30.5  m
(100  ft)  and averaging about 10 m.  Within these  limits  the
hydraulic  gradients were estimated, checking the thickness  of
the   surficial  aquifer  and  confining  layer  to  ensure   a
physically realistic system.

2.2.5.1.3   Porosity—Porosity  is the percent void space in  a
representative  volume of porous media.  Below the water  table
in  the  ground water all the void space is filled with  water.
As   shown   in   equation  2.1,  the  porosity   affects   the
ground-water   velocity.   Given  that  all  other   hydrologic
properties  are the same, a low porosity will result in  higher
ground-water  velocities.  This happens because there are fewer
spaces  for  the water to flow through so the water  must  flow
faster for a given specific discharge.

The  limestone  geology  of the Floridan peninsula  results  in
extremely   variable  porosities.   In  areas  where   solution
cavities  and  channels are common, porosities are easily  over
.50.   Other areas have layers of dolomite and chert which  are
more  resistant  to dissolution where the porosities can be  as
low  as   .05.   This  makes  estimation  of  the  bulk  aquifer
porosity  very  difficult.   A  lab determined  porosity  of  a
massive  limestone  can not accurately represent  the  solution
features  that are present.  The modeling studies reviewed  did
not   consider  transport  so  porosity  was  not  a  necessary
parameter.   Stewart (1966) reported a range of porosities from
.15  to   .44  calculated in the lab for Polk  County.   Tibbals
(1984,  personal  communication) suggested bulk  porosities  in
the  Floridan  Aquifer  would  probably  be  about  .3  to  .35
                             116

-------
although   not   at  all  uniformly  distributed.   Some   U.S.
Geological  Survey modeling studies have used a porosity of  .2
for  the  Floridan  with good success (Frazee,  1984,  personal
communication).    Solution  features are most highly  developed
in  the northwest area where the Floridan is unconfined.  Based
on  data available the porosities shown on Table 2.13 were used
for  modeling.   Because  this  parameter  was  not  very  well
defined,  some sensitivity runs were done varying the  porosity
values.   Porosities  of .15 to .20 for the  surficial  aquifer
and  .05  to  .10 for the confining layer  of  the  two-aquifer
system  were  determined  in a similar manner  (Tibbals,  1984,
personal  communication; Hutchinson, 1978; Cherry et al., 1970;
and  Knochenmus and Hughes, 1976).  The worst case porosity  is
less  than the average porosity because the lower porosity will
increase  the rate of ground-water movement as discussed at the
beginning of this section.

2.2.5.1.4   Saturated Thickness—When modeling the transport of
aldicarb  through  an aquifer to a well the critical  depth  is
the  depth  of  the  well rather than the  full  depth  of  the
aquifer.   Parts  of the Floridan Aquifer are over 600 m  thick
(Causey  and  Leve, 1976).  When a municipal supply  well  only
penetrates  the  upper 100 m it is not critical that the  exact
aquifer thickness be known or modeled.

When  modeling  a multi-aquifer system, however, the  thickness
of  the upper unconfined aquifer and the intervening  confining
layer  are critical factors determining the speed of  transport
to  a  well  pumping out of the lower aquifer.   The  following
analysis   demonstrates   the  importance  of   the   saturated
thickness  to the average vertical velocity of the water.   The
travel  time through the upper aquifer and confining layer to a
well  is  equal  to  the  quotient  of  the  vertical  distance
traveled  and  the velocity.  By substituting Darcy's  Law  for
velocity it is shown that:

                             117

-------
          t =   i                                        (2.2)
              KAH
where     b  = saturated thickness
          t  = travel time
          e  = effective porosity
          K  = hydraulic conductivity and
          AH = difference in hydraulic heads between the upper
               and lower aquifers

This  analysis  shows that the saturated thickness (b)  is  the
most  important factor in determining the travel time, since it
is the only squaired term.

In  Area  1,  where  the Floridan  Aquifer  is  unconfined,  an
average  saturated  thickness of 183 m (600 ft) was  determined
from  Ryder  (1981)  and Causey and Leve (1976).   In  Area  4,
where  the  unconfined surficial aquifer is  considered  alone,
the  average saturated thickness of 23 m (75 ft) was determined
from  values  sited  in  county reports pertaining  to  Area  4
(Wedderburn  et  al.,  1982; Wolansky,  1978;  Lichtler,   1960;
Lichtler,  1972; and Bearden, 1972).  The same values were used
for  both the average and the worst cases because the saturated
thickness  is  not a critical value in the case of  the  single
aquifer as discussed above.

The   same  procedure  was  used  to  determine  the  saturated
thickness  for the two-aquifer system (Areas 2, 3 and 5).   The
surficial  aquifer  is extremely variable in thickness  in  the
central  highlands.  In Polk County it can vary from less  than
15  m   (50 ft) to as much as 75 m (250 ft)  (Stewart,  1966  and
Hutchinson,  1978).   The average saturated thickness is   23  m
(75  ft), the same as for the surficial aquifer in the  coastal

                             118

-------
area.   The  vertical  travel time decreases as  the  saturated
thickness  decreases as shown in the analysis above.  The worst
case value is 12 m (40 ft).

The  confining  layer is also extremely variable in  thickness.
In  areas  where  the clays are breached by a  sink  hole,  the
thickness  is  effectively zero.  Knochenmus and Hughes  (1976)
report  that  in  parts  of Lake County the  thickness  of  the
confining  layer reaches as much as 30.5 m (100 ft).  Buono  et
al.  (1979) map the thickness of the confining bed in parts  of
this  area  as less than 7.5 m  (25 ft) to over 75 m  (250  ft).
The  average thickness is 12 m  (40 ft) and the worst case is  3
m  (10 ft).  Where the confining layer is actually breached the
unconfined   Floridan  configuration  can  be  used  to  assess
pesticide exposure.

The  Floridan Aquifer is estimated to be 180 m (600 ft) in  the
central  area.   As  in  the case of the  single  aquifer,  the
saturated  thickness is not as  critical a value as the depth of
a drinking water supply well.

2.2.5.1.5   Recharge—Recharge  is the average annual  rate  at
which  water  is  replenished   to the  aquifer  by  percolation
through  the  urisaturated  zone.  The amount of water  that  is
recharged  to the ground water  can be roughly figured by a mass
balance analysis, where:

     Recharge = Precipitation + Irrigation - Runoff
                - Evapotranspiration

Other  factors  influence  the  recharge rate as  well  as  the
supply   (precipitation  and irrigation).  In areas where  soils
are  poorly drained recharge rates are close to zero.  The well
drained   soils  typical  of  the  ridge  area  (entisols   and
ultisols)   enhance   recharge   rates.   Lakes   and   natural

                             119

-------
depressions   such  as  sink  holes  create  ideal  areas   for
recharge.   Another  factor that affects the recharge  rate  is
the  height of the water table.  If the water table is near the
land   surface,  extra  water  added  to  the  aquifer  can  be
evaporated more easily.

Stewart  (1980) delineates areas of high, moderate, low and  no
recharge  to  the Floridan Aquifer.  This map gives  a  general
picture  of  recharge  both where the  Floridan  is  unconfined
(Area  1) and where it is confined (Areas 2, 3, and 5).  On the
average,  recharge is higher where the Floridan is  unconfined.
In  the  central  portion of the peninsula where  the  Floridan
Aquifer  is confined there are areas of high recharge along the
ridges,  but  much  of the area has low to  moderate  recharge.
Estimates  of  recharge  from the map were  compared  to  rates
reported  in  the  literature.  Ryder (1982)  reports  recharge
rates  that  range  from 0 to 50 cm/yr (20 in/yr) in  the  area
where  the  Floridan  is  unconfined.  Where  the  Floridan  is
confined  recharge is reported anywhere from about 7 cm/yr (2.6
in/yr)   (Hutchinson, 1978) to 38 cm/yr (15 in) (Knochenmus  and
Hughes,  1976).  The rates used in the modeling study are shown
in  Table  2.13.  Both the average and the worst case  recharge
rates   also  reflect  the  effects  of  added  recharge   from
irrigation.

Recharge  to  the surficial aquifer has not been  well  studied
where  the  Floridan is not considered.  A rough  mass  balance
approach  was used to estimate the recharge rates.  Visher  and
Hughs   (1969)  have  published  a map  showing  the  difference
between  rainfall  and  potential evaporation.  This  was  used
along  with  estimates of runoff (Hughes, 1976) and  irrigation
to  calculate the average and worst case recharge rates for the
surficial aquifer.
                             120

-------
2.2.5.2  Drinking Water Well Parameters—

For  the purposes of modeling well depths and pumping rates are
necessary.    Water  supply  wells  of  different  depths  were
modeled  in  each hydrogeologic setting based on typical  wells
in  each area.  Where the Floridan Aquifer is unconfined both a
deep  well  and  a  shallow  well  were  simulated.   The  deep
municipal  wells  in the Floridan Aquifer typically draw  water
from  about 110 m (350 ft) of well which is open to the aquifer
(Tibbals,  1984,  personal  communication and  Spangler,  1984,
personal  communication) .   The  shallov/ wells vary  from  less
than  a  meter up to about 30 m (100 ft), with an average 12  m
(40 ft).

In  the case where there is just a single unconfined  surficial
aquifer,  only  a shallow well was modeled.  Well  depths  vary
considerably  in  the surficial aquifer.  An average  depth  of
10.7 m (35 ft) was used in the simulation.

Since   the  two-aquifer  system  is  modeled  to  assess   the
possibility  of  the  pesticide being transported  through  the
confining  layer  to the confined aquifer, only a deep well  in
the  lower  aquifer  was simulated.  Wells tapping  a  confined
aquifer  are  typically open in the highest producing zone  and
have about 100 m (350 ft) open in the aquifer.

The  pumping  rates  modeled with each  hydrogeologic  scenario
reflect  the two most common drinking water supply  situations:
a  small   (low yielding), domestic supply well and  the  large,
high  volume, municipal supply well.  Representative rates were
chosen  based  on well yields and annual water usage  from  the
communities and well records in the citrus growing area.

The  largest  capacity  wells  are  in  the  confined  Floridan
Aquifer.   Large  municipal supply wells have rates as high  as

                             121

-------
94.6  1/s  (1500  gpm).  Deep wells in  the  confined  Floridan
Aquifer  also  have  high  rates.  The  shallow  wells  in  the
surficial  aquifer  have much lower rates ranging from 2.5  1/s
(40  pgm) to 12.6 (200 gpm).  Table 2.14 shows the well  depths
and rates used with each hydrogeologic setting.
                              122

-------
 TABLE 2.14  WELL RATES AND DEPTHS FOR SATURATED ZONE MODELING
Aquifer Configuration

Surficial
Floridan unconfined
Two-aquifer system
(1 ft  = .3048 ra)
(1 gpm = 0.063 1/s)
Well Depth

  35 ft
  40 ft
                            350 ft
 350 ft (open)
Well Pumping Rate

    40 gpm

   200 gpm


    40 gpm

   200 gpm

   500 gpm

  1000 gpm


   700 gpm

  1500 gpm
                              123

-------
                           SECTION 3
                  CHEMICAL FATE AND TRANSPORT
Aldicarb   is   a   nematicide,  acaricide,  and   a   systemic
insecticide.   Its  environmental  fate  is  dominated  by  two
factors:  the  fact that it forms two toxic daughter  products,
and  its  high  mobility in soils.  Degradation  of  the  toxic
residues  of the compound is of intermediate duration  compared
to other pesticides.

This  discussion  will be broken into three  primary  sections;
the  first  dealing  with factors which  affect  aldicarb  fate
(i.e.,  transformation  and  decay), the  second  dealing  with
factors   affecting   transport  (i.e.,  adsorption   partition
coefficients)  and the third combining those two factors in the
analysis  of  pesticide  parameters  for  the  unsaturated  and
saturated zone.
3.1  ALDICARB FATE
Aldicarb  is  a white crystalline solid which  is  incorporated
into  soil  as  a granule containing either 10% or  15%  active
ingredient.   In  order  to be effective, it must  dissolve  in

                              124

-------
water.   Once  this  happens in soils, the compound  begins  to
transform and degrade.

The  current  theory is that aldicarb,  2-methyl-2-(methylthio)
propionaldehyde  0-(methylcarbamoyl)  oxime, is fairly  rapidly
oxidized  to  aldicarb  sulfoxide,  2-methyl-2-(methylsulfinyl)
propionaldehyde  0-(methylcarbamoyl)  oxime  which in  turn  is
more   slowly   oxidized  to  aldicarb   sulfone,   2-methyl-2-
(methylsulfonyl)   propionaldehyde  0-(methylcarbamoyl)  oxime.
Concurrently,   these  three  carbamates  are  transformed   by
hydrolysis  to corresponding oximes.  Hydrolysis is a  chemical
reaction  in  which water breaks up an organic  molecule  (RX),
such  as aldicarb, by breaking a carbon-X bond and replacing it
with OH  from the water molecule:
              R-X + H0—*>R-OH + X~+ H+                 (3.1)
These  products of hydrolysis are far less toxic than aldicarb,
its  sulfoxide  or its sulfone and are of little  environmental
concern   (Smelt et al., 1978a).  A schematic of these processes
is shown  in Figure 3.1.
3.1.1  Transformation and Degradation Rates in Soils
A  number  of  studies  up until 1978  were  performed  on  the
pesticide   aldicarb  as  reported  by  Smelt  et  al.,  1978c.
Unfortunately,  most of these studies only measured the residue
remaining  in  soil after application and therefore  cannot  be
used  to  identify  rates  for  individual  transformation  and
degradation  processes.  Smelt et al., 1978c, identified  these
field   loss  rates for aldicarb being, from 0.015 to  0.42/day.

                             125

-------
                 Aldicarb
      Aldicarb Sulfoxide
                                                          Aldicarb Sulfone
CH.S - C - CH - NOCNHCH.
  3    |
    O  CH0
    II   I 3
•CH3S - C  - CH

       CH.,
                      II            II    I
                  = NOCNHCH.,— -CH,S -  C -  CH  =
                                                                   'II
                                       I
                                      CH.
                                                                                  O
                                                                                  ||
                                                                                NOCNHCH,
                             (Hydrolysis)
                        (Hydrolysis)
to
Nontoxic Oximes and  Nitrites
               Figure 3.1  Schematic of aldicarb environmental  chemical Pathways

-------
They   also   performed  independent  experiments  to   measure
oxidation  and hydrolysis rate constants for the pathways shown
in  Figure  3.1.   These computed rate constants are  shown  in
Table  3.1.  Other workers (Bromilow et al., 1980 and  Leistra,
et  al.,  1976)  have performed similar experiments  which  are
also shown in Table 3.1.
3.1.2  Hydrolysis Rates in Water
Hansen  and  Spiegel  (1983) measured the hydrolysis  rates  of
aldicarb,  aldicarb sulfoxide and aldicarb sulfone in distilled
water  at  various levels of temperature and pH.  These  values
are  shown  in Table 3.2, taken from their report.  Lemley  and
Zhong  (1983)  measured acid and base catalyzed  hydrolysis  of
aldicarb,  aldicarb  sulfoxide  and aldicarb  sulfone.   Second
order  rate  constants   were calculated for data  taken  under
extremely  acidic  and  basic conditions.  Rate  constants  for
base  catalyzed hydrolysis of aldicarb, aldicarb sulfoxide  and
aldicarb  sulfone were 0.94 liter/mole-min, 10.5 liter/mole-min
and   40.3  liter/mole-min,  respectively.   At  pH  7.5  these
correspond   to   first-order  rates  of  0.0004,  0.0048   and
0.0184/day.   These  rates agree well with the data  of  Hansen
and  Spiegel  for aldicarb but are roughly 4 times greater  for
the  sulfoxide  and  3 to 10 times greater  for  the  sulfone.
They  also reported strong temperature effects   (14 fold over a
5  deg.  C to 35 deg. C range) on base catalyzed hydrolysis  of
the  sulfone and a 'depression of hydrolysis in the presence  of
neutral  electrolytes.   However, in the presence of 1 m  NaCl,
the rates at 15 deg. C were not quite halved.
                              127

-------
          TABLE  3.1
FIRST ORDER RATE  CONSTANTS FOR OXIDATION AND HYDROLYSIS  OF

ALDICARB IN SOIL
to
00
SOIL
Clay Loam
(fresh)
Clay Loam
(stored)
Greenhouse
(fresh)
Greenhouse
(stored)
Peaty Sand
(fresh)
Peaty Sand
(stored)
Sandy Loam
Clay Loam
Clay Loam
Greenhouse
Soil
Greenhouse
Soil
Sandy Loam (3)




Sandy Loam (5)




Nagele. Loam
Westmaas Loam
Ki
0.30

0.30

0.096

0.160

0.130

0.077

0.210






0.30
0.44
0.21
0.80
0.80
0.20
0.27
0.14
0.46
0.55
0.25
0.25
RATE CONSTANT (/day)
KJ K| K^
0.011

0.008

0.004

0.010

0.008

0.006

0.006
0.005
0.020
0.0036

0.027

0.015
0.033
0.034
0.035
0.025
0.011
0.030
0.013
0.031
0.031
0.015-0.03
;o. oi-o. 06
0.015

0.030

0.000

0.000

0.002

0.0007

0.005






0.05
0.07
0.06
0.07
0.07
0.04
0.065
0.07
0.07
0.07
0
0
0.013

0.018

0.0015

0.005

0.006

0.005

0.008
0.004
0.030
0.0016

0.013

0.003
0.010
0.006
0.018
0.021
0.001
0.004
0.002
0.006
0.01
.00-0.02
.00-0.05
K5
0.03

0.021

0.004

0.013

0.0045

0.0045

0.012
0.008
0.050
0.004

0.033

0.012
0.020
0.013
0.021
0.016
0.005
0.010
0.005
0.012
0.015
0.00-0.08
0.00-0.14
TEMP
•c
15

15

15

15

15

15

15
6
25
6

25

5
10
15
IS
15
5
10
15
15
15
20
20
PH
(su)
7.2

7.2

6.1

6.1

5.1

5.1

7.4
7.2
7.2
6.1

6.1

7.0
7.0
7.0
7.0
7.0
6.3
6.3
6.3
6.3
6.3
7.2-7.4
7.3-7.6
ORGANIC
MATTER
(*)
4.5

4.5

17.2

17.2

9.5

9.5

1.7
4.5
4.5
12.2

17.2

1.35
1.35
1.35
1.35
1.35
5.92
5.92
5.92
5.92
5.92
2.5-4.6
1.1-4.7
HATER
CONTENT
(%)
30.0

30.0

66.0

66.0

19.0

19.0

13.5
30.0
30.0
66.0

66.0

10.0
J.0.0
5.0
10.0
15.0
10.0
10.0
5.0
10.0
15.0


REFERENCE
Smelt et al.

M mm

m mm

m mm

m mm

m mm

m mm
Smelt et al.
• mm
m mm

m mm

Bromilow et al
m
m
m
m
m
"
m
m
m
Leistra et al.
• mm

1978c

m

m

m

m

*

H
1978b
•
II

•

. 1980









1976
•

-------
    TABLE 3. 2
FIRST ORDER HYDROLYSIS  RATE  CONSTANTS FOR ALDICARB; ALDICARB SULFOXIDE

AND ALDICARB SULFONE
to
«£>
COMPOUND
Aldicarb





Sulfoxide





Sulfone





pH
5.5
5.5
7.5
7.5
8.5
8.5
5.5
5.5
7.5
7.5
8.5
8.5
5.5
5.5
7.5
7.5
8.5
8.5
TEMP (°C)
5
15
5
15
5
15
5
15
5
15
5
15
5
15
5
15
5
15
PERIOD
(days
1-186)
1=186
1-186
1-186
1-186
1-186
1-186
1-278
1-278
1-186
1-186
1-186
1-83
1-186
1-186
1-186
1^186
1-186
1-28
RATE CONSTANT
(d-«)
0.000151
0.000214
0.000354
0.000363
0.000501
0.00407
0.000371
0.00158
0.00107
0.00191
0.0107
0.063
0.000741
0.00155
0.00110
0.00550
0.0200
0.134
HALF-LIFE
(d)
4,580
3,240
1,950
1,900
1,380
170
800
440
650
360
65
10
930
450
630
125
35
5
    Source:   Hansen and Spiegel, 1983

-------
3.2  TRANSPORT CONSIDERATIONS
3.2.1  Adsorption
Aldicarb  and its two daughter products are very water soluble,
which  indicates  a  high mobility in soils.   Cohen   et  al.,
(1983)  reported  solubilities of 6,000, 43,000 and  7,800  ppm
for  aldicarb,  sulfoxide and sulfone,  respectively,  although
Hornsby  et  al.,  (1984)  give slightly  different  values  of
6,000, 33,000 and 8,000 ppm.

Using  these solubilities and the regression equation of Kenaga
and  Goring (Lyman, 1982), Koc values of 36 cm /g for aldicarb,
12  to  14 cm /g for aldicarb sulfoxide and 31 to 32 cm /g  for
aldicarb sulfone are computed.

Based  on  the data of Bromilow et al.  (1980) (reported as K,  )
for  a sandy loam soil and a sandy loam soil with  addition  of
peat,   Koc  for  aldicarb,  aldicarb   sulfoxide  and  aldicarb
                          33                     3
sulfone  were  16-70  cm  /g,  4.6 cm /g and  1  to  5  cm  /g,
respectively.  Koc is calculated by:
                          K   =                          0.2)
                           °C  (OC)
where     K  =  the adsorption partition coefficient (cm /g)
          OC =  the fraction organic carbon

Hornsby  et  al.  (1984) calculated values from 0 to 47  cm   /g
for  aldicarb  and  0 to 18 cm  /g for aldicarb  sulfone  at  a
Florida  ridge citrus site and 0 to 38 cm /g for aldicarb   and
0  to 11 cm /g for aldicarb sulfone at a flatwoods citrus site.

                             130

-------
They  also  reported   Koc  calculated for  other  soils  which
ranged  from  19
aldicarb sulfone.
ranged  from  19  to  25  for aldicarb and 0 to  6  cm  /g  for
The  various  values of Koc collected from the  literature  are
summarized in Table 3.3.

Note  that  Hough et al. (1975) found  adsorption  coefficients
for  aldicarb sulfoxide to be an order of magnitude higher in a
soil  with high clay content but having the same organic matter
content as another soil with a lower coefficient.

Supak  (1972), on the other hand, found aldicarb to actually be
excluded  by Ca- and Al- saturated montmorillonite clays.   Ca-
and  Al-  saturated  illite and kaolinite clays  showed  weakly
positive  adsorption isotherms and adsorption was thought to be
through interaction with water on the external clay surface.
3.2.2  Volatilization
Because  aldicarb  is  applied in a  granular  formulation  and
incorporated  it does not seem that volatilization would be  an
important  process.   However,  it does have a  moderate  vapor
pressure  (1x10 ~^ mm  Hg  @  25  deg.  C)  compared  to  other
pesticides.    Henry's  Law  Constant  calculated  using   this
information  (Thibodeaux,  1979) gives  a value of 1.7 x  10
   3       /  3
cm  -water/cm -air.  This value is comparable to that of  other
pesticides such as DDT, dieldrin, trifluralin and EPTC.

Bull  et al. (1970) determined the effects of soil moisture and
temperature  on aldicarb volatilization from vials containing a
treated  layer of sand capped by 5.5 cm of untreated sand.   In
their  experiments,  volatile  losses of aldicarb  ranged  from

                             131

-------
     TABLE 3.3  KQC FOR ALDICARB AND ITS DAUGHTER PRODUCTS
                                    K   (cm3/g)
                                     oc

                        Aldicarb	Sulfoxide	Sulfone

Calculated from
Solubility data            36            12-14          31-32
Bromilow et al.f 1980
   Woburn (3)              70              -              1
   Woburn (5)              16              5              5
Hornsby et al.,  1984
   Lake Hamilton site     0-47             -             0-18
   Oviedo site            0-38             -             0-11
   Eustis soil           19-25             -             0-6
   Webster soil            21              -              6
   Cecil soil              19-0
   Grenada soil            19              -
Hough et al., 1975
   Holtville clay          -              236
   Buren silt loam         -               24
                             132

-------
16.7%  to 89.0% over a 24 hr. period.  Temperatures ranged from
25  deg. C to 75 deg. C.  Losses were generally higher from wet
than  from dry sand.  First-order volatilization rate constants
computed  over  the 24 hr. period were 0.0036 and  0.0076  /day
for  the  wet and dry sand at 25 deg. C and 0.0347  and  0.0180
/day for wet and dry sand at 50 deg. C.

Other  evidence  presented  in  Supak (1972),  as  reported  in
INTERA   (1980),  indicates  that  volatilization  losses   are
extremely  low.   Supak  et al. (1977),  report  extremely  low
losses  (0.01  to  0.18%) over an 18 day period  in  two  Texas
soils.   They  also  found volatilization losses to  be  higher
from  a  dry soil than from a wet soil although  volatilization
ceased under very dry conditions.

Richey  et  al.r 1977, radio-labeled aldicarb at the  S-methyl,
N-methyl   and   tertiary  carbon  atoms  to  study   molecular
fragmentation  under soil degradation.  As a result, they found
virtually  all of the radio activity captured as volatiles  was
in  the form of CO  and less than 1% captured as volatiles  was
in  the  form  of  aldicarb,  aldicarb  sulfoxide  or  aldicarb
sulfone.   Thus,  based on this study, it appears  that  little
volatilization of the parent and major daughters would occur.
3.2.3  Plant Uptake
Some  plant  uptake  literature  was reviewed  by  Pacenka  and
Porter   (1981).   Their  investigations  revealed  that  little
quantitative  data is available.  Based on limited information,
they  estimated that 4 to 20% of the applied chemical might  be
a   suitable   range.    Biggs   and   Webb   (1984,   personal
communication)  have  data for the uptake of aldicarb by  young
Valencia  orange trees.  A summary of their data indicates that

                             133

-------
as  much as 34% of the applied chemical was taken up by  plants
by   30   days  after  application  (Biggs  and  Webb,   1983).
Approximately  20%  of applied aldicarb was calculated to  have
been  taken  up  by 20 year old trees in the field  (Biggs  and
Webb, 1984, personal communication).
3.3  ANALYSIS OF PESTICIDE MODEL PARAMETERS
From  the  information  identified  from  the  literature,  the
parameters   needed  for  simulation  of  pesticide  fate   and
transport  using PRZM and CFEST were derived.  In the following
analyses  the decay rates of. aldicarb and its daughter products
are  estimated in the unsaturated and saturated zones based  on
the  appropriate  physical properties.  In the  saturated  zone
the  pH  and  temperature  of the ground  water  are  the  most
important  factors  influencing the pesticide degradation.   In
the  unsaturated  zone these factors are important as  well  as
soil water content and amount of organic carbon.
3.3.1  Unsaturated Zone
Three  different data sets  (Bromilow et al., 1980; and Smelt et
al.,  1978b  and  1978c) were identified  in  which  laboratory
determined  rates of aldicarb conversion to the sulfoxide  (k^),
the  conversion  of the sulfoxide to the sulfone  (k  ), and  the
degradation  of each species  (k  , k  and k  ) to nontoxics  were
                               34      5
estimated.   In  these studies, measurements were also made  of
levels  of  temperature, pH, organic matter content  and  water
content   at  which  the  experiments  were  conducted.   Eight
different  soils  were used having a pH range from 5.1 to  7.4.
Temperatures  ranged  from  5 to  25 deg. C, organic matter  from

                            . 134

-------
1.35  to  17.2%  and water contents from 5 to  66%.   Seventeen
meaurements  of k, and k , were made versus these  environmental
variables,  while twenty-one measurements of k~, k. and k5 were
made.
The   data  were  analyzed  with  multiple  linear   regression
techniques.   It was felt that the major factors affecting  the
constants  k,  and  k ~ (since they involve an  oxidation  step)
would  be temperature,  water content and organic matter (if the
reaction   is   microbially  mediated).    The   major   factors
affecting  k  , k  and k  would be temperature, pH and  perhaps
water   content  since  these  reactions  are  thought  to   be
principally hydrolysis.

The  results  of the regressions are shown in Table  3.4.   The
percent  variance  in  k  and k , explained  by  the  variables
temperature,  OM  (organic matter) and WC (water  content),  is
low.   Stepwise regression indicates that most of the  variance
is  explained by organic matter (29%) followed by water content
(11%).   Temperature explains only about 1%.  Regressions  were
also  performed for k  and k  subsituting pH for water content.
                                                 2
These  regressions  had slightly lower multiple R   than  those
shown in Table 3.4.
Regression  of  k  ,  k  and k  on temperature,  pH  and  water
                 J     T      J                       O
content  produced  more favorable results.  Multiple R   values
were  0.50, 0.65 and 0.71, respectively.  Temperature explained
the  largest  percentage of the variance in the  k_  regression
(46%)  followed by pH (21%).  Water content had little  impact,
explaining only about 1% of the variance.

The  linear regression of these rate constants on environmental
variables  provided  a means of evaluating pesticide  parameter
values   for   different  unsaturated  zone   scenarios.    The
equations  are  purely  empirical, however.   Nicholls  et  al.

                             135

-------
    TABLE 3.4  MULTIPLE LINEAR REGRESSION COEFFICIENTS FOR

               ALDICARB TRANSFORMATION AND DEGRADATION RATES
               	Regression Coefficients
         R'    Inter-      Temp      pH       OM       WC
               cept        (°C)               (%)      (%)
Rate
 k      0.38    0.30       0.01      ~      -0.04    4.36E-3


 k      0.31    0.01      7.6E-4     —      -1.8E-4  -2.2E-4
  2

 k      0.50   -0.01     -5.6E-5    0.01        —   -9.99E-4


 k4     0.65   -0.04      9.6E-4    0.01        —    -3.0E-5


 k      0.71   -0.06      1.4E-3    0.01        —     3.9E-5
                             136

-------
(1982)  used  a  semi-empirical relationship to  correlate  the
first-order   rates   with   environmental   variables.    That
relationship was:
            ln(k) =  a + £ln(wc) + y(l/T)                 (3.3)
which  uses  only  the  variables (we) water  content  and  (T)
temperature.

They  found values of  a = 26.5,  3= 0.407 and  y= -8571 for the
aldicarb  sulfone  degradation  rate  (i.e., kj.)  in  a  Compton
Beauchamp   soil.    The  present  analysis  using   the   same
regression  equation  for eight different soils gave  a=  17.8,
 g  =  0.04  and y = -6349 for aldicarb sulfone.   However,  the
                                                           2
regression  coefficient  of  determination was very low   (R   =
0.24).   This  was not  unexpected since pH was found to  be  an
important regression variable for rate k5.

Lacking  a  better  data set which is  more  representative  of
Florida   soils,  or  a  better  (more  theoretical)  means  of
analyzing  the experimental rates, the above data and  multiple
linear  regression approach were used to calculate  first-order
transformation  and  degradation  rates for  aldicarb  and  its
daughter products for each unsaturated zone scenario.

Table  3.5  shows characteristic information used  to  estimate
model  pesticide  parameters by soil  and by horizon.   Although
soil  pH  and % organic carbon are not specifically  inputs  to
PRZM,  they were used to calculate pesticide degradation  rates
and   adsorption   partition  coefficients  which  were   input
directly. These rates and coefficients are shown in Table 3.6.
                              137

-------
TABLE 3.5  CHARACTERISTIC DATA FOR FLORIDA SOILS USED TO
           ESTIMATE PESTICIDE PARAMETERS BY HORIZON
Soil
Group
Entisols
Entisols
Entisols
Entisols
Entisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Alfisols
Alfisols
Alfisols
Alfisols
Alfisols
Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
Spodosols
Horizon
1
2
3
4
5
1
2
3
4
5
6
7
1
2
3
4
5
1
2
3
4
5
6
Depth
(cm)
0-10
10-20
20-40
40-100
100+
0-20
20-40
40-100
100-160
160-200
200-240
240+
0-20
20-40
40-80
80-120
120-240
0-10
10-40
40-60
60-100
100-200
200+
efc
(%)
5.6
5.4
4.2
3.5
3.0
8.2
7.0
5.5
8.0
20.0
22.5
5.0
10.0
6.5
9.5
20.0
25.0
14.0
9.0
6.0
12.0
21.0
6.0
PH
(su)
5.2
5.3
5.5
5.6
5.7
5.8
5.8
5.7
5.4
5.2
5.1
5.0
5.8
6.1
6.7
6.2
6.5
4.5
4.7
5.2
5.2
5.5
5.6
OC
(%)
0.8
0.7
0.4
0.15
0.06
0.8
0.65
0.25
0.10
0.05
0.05
0.05
1.2
0.4
0.1
0.2
0.2
2.1
1.2
0.35
0.90
0.90
0.40
PS 3
(g/cm )
1.35
1.4
1.45
1.45
1.5
1.5
1.52
1.54
1.56
1.64
1.66
1.68
1.4
1.5
1.6
1.65
1.68
1.3
1.45
1.50
1.50
1.50
1.50
                           138

-------
    TABLE  3.6
TRANSFORMATION AND  DEGRADATION  RATES  (k) AND ADSORPTION
PARTITION COEFFICIENT (K) USED  IN MODELING  ALDICARB FATE
AND TRANSPORT IN FLORIDA SOILS
     Soil Group  Horizon
         Depth
     Entisols
    Ultisols
vo
    Alfisols
    Spodosols
  1
  2
  3
  4
  5
  1
  2
  3
  4
  5
  6
  7
  1
  2
  3
  4
  5
0-10
10-20
20-40
40-100
100+
0-20
20-40
40-100
100-160
160-200
200-240
240+
0-20
20-40
40-80
80-120
120-240
0-10
10-40
40-60
60-100
100-200
200+
0.52
0.53
0.54
0.56
0.56
0.53
0.54
0.56
0.58
0.63
0.64
0.57
0.51
0.55
0.58
a. 62
0.65
0.47
0.51
0.55
0.54
0.58
0.55
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.02
0.03
0.03
0.03
0.03
0.02
0.02
0.03
0.03
0.03
0.03
0.02
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.05
0.03
0.04
0.02
0.03
0.03
0.03
0.02
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.04
0.05
0.05
0.05
0.03
0.03
0.04
0.04'
0.04
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.04
0.04
0.04
0.04
0.02
0.02
0.03
0.03
0.03
0.03
0.16
0.14
0.08
0.03
0.01
0.16
0.13
0.05
0.02
0.01
0.01
0.01
0.24
0.08
0.02
0.04
0.04
0.42
0.24
0.07
0.18
0.18
0.08
0.11
0.10
0.06
0.02
0.01
0.11
0.09
0.04
0.01
0.01
0.01
0.01
0.17
0.06
0.01
0.03
0.03
0.29
0.17
0.05
0.13
0.13
0.06
0.07
0.06
0.04
0.01
0.01
0.07
0.06
0.02
0.01
0.004
0.004
0.004
0.11
0.04
0.01
0.02
0.02
0.19
0.11
0.03
0.08
0.08
0.04

-------
Rates  k, and k« were determined by entering soil  temperature,
organic  matter  and  water content data  into  the  regression
equations shown in Table 3.4.

An  average  soil  temperature of 25 deg. C was  used  for  all
layers/  in all soils.  Soil temperature data from Gainesville,
Lake  Alfred  and Bradenton were used to draw this  conclusion.
The  average soil temperature at Gainesville (4 and 8"  depths)
for  1980,  1981 and 1982 ranged between 23 and 24 deg.  C  (73
and  75 deg. F).  Average temperatures at the 4" depth at  Lake
Alfred  and  Bradenton over the same three year period were  26
and  24 deg. C, respectively.  Since ground-water  temperatures
seem  also  to average at about this level, a 25 deg.  C  level
seemed appropriate for the soil.

Organic  matter was assumed to be 1.7 x organic carbon and  the
water  content of a horizon was taken to be its field  capacity
value.

Rates  k_, k. and k- were determined by entering  temperatures,
pH   and   water  contents  into  the  appropriate   regression
equations in Table 3.4.

The  adsorption  partition  coefficients  K , K-  and  K3  were
determined  by  assuming Koc for aldicarb,  aldicarb  sulfoxide
and  aldicarb  sulfone to be 20, 14 and 9 cm  /g  respectively,
and  multiplying by the appropriate organic carbon content from
Table 3.5.

Dispersion/diffusion  coefficients  were  set to zero  for  all
horizons.   Plant uptake efficiency factors were set to one  in
horizons within the root zone and zero elsewhere.

An  application of 5.61 kg/ha (5 Ib/ac) was made on February 15
of   each   simulation  year.   This  is  the   current   label

                             140

-------
application  rate  for  Florida citrus.  The  application  rate
corresponds  to the optimal date for pest control (Jones, 1984,
personal  communication).   The  pesticide was  assumed  to  be
incorporated to a five centimeter depth.
3.3.2  Saturated Zone
The  fate and transport of aldicarb and its metabolites in  the
saturated  zone differ significantly from the unsaturated zone.
Factors  such  as  volitalization and plant  uptake,  that  are
important  in the unsaturated zone, do not influence  transport
in   the  saturated  zone.   Adsorption  of  aldicarb   depends
primarily  on  the amount of organic carbon available  to  bind
the  compound  to  the  solid phase,  although  there  is  some
conflicting   evidence  of  the  influence  of  clays  on  this
process.   Because there is an insignificant amount of  organic
carbon   in  the  saturated  zone,  adsorption  is   considered
negligible.

The  fate  of aldicarb seems to be dominated by  hydrolysis  in
the  saturated zone.  There are four key factors that influence
the  rate  of  degradation in the ground water in  the  Florida
citrus growing area.

    1)   temperature of ground water
    2)   pH of ground water
    3)   ionic strength
    4)   physical and chemical characteristics of the aquifer
         material

The  effects  of  the ionic strength on kinetic  reactions  are
complex  and poorly understood, and therefore also difficult to
include  in  a quantitative analysis of the  degradation  rate.

                             141

-------
•The   following  discussion  qualitatively reviews  the  current
 thinking  on  this   factor  and discusses   its  effect  on  the
 hydrolysis rate..

 The   ionic  strength  (I)  of a solution is a  measure  of  the
 influence of  the  free  ions:
                       n      2
                   I =  L Ci^i                             (3.4)
 where      I   =  ionic  strength
           C.  =  concentration of  ion  i
           Z .  =  valance of  ion  i

 Because   the  hydrolysis reaction  includes  ionic   species,   the
 hydrolysis rate  is depressed  in the presence  of  ions  competing
 for  the   same  reaction sites as  the  OH ion  of   water.    When
 hydrolysis rates in distilled water with no competing   ions
 were   compared  to rates in well water for   aldicarb  sulfoxide
 and  aldicarb  sulfone   (Lemley  and  Zhong,  1984),   the  results
 showed the   rates in well water were  almost half the  rates  in
 distilled water.  The  ionic strength of  the well  water  is
 assumed   to   be an important factor  causing the rate  decrease.
 Although   the  actual ionic strength of the well  water was   not
 stated,   an   earlier  study (Lemley and Zhong,  1983)  tested   the
 effect of competing  ions  directly by  comparing the hydrolysis
 rate   in   a   NaCl solution with  an   ionic  strength  of   1.0
 equivalant/liter   to  hydrolysis  in distilled water for aldicarb
 sulfone.   The  rate  was   36%  slower  in   the   1  Molar   Had
 solution.

 The    ionic   strength of   the  ground  water   in  Florida   was

                              142

-------
approximated  from  water  quality data (Phelps, 1978)  in  the
major   citrus  growing  counties  to  assess  the  impact   of
competing  ions  on  the hydrolysis rate.  The  ionic  strength
ranges  from .01 to .04 equivalents per liter as compared to  1
equivalent/liter  in the study cited above.  Based on the rough
estimates  of ionic strength of the ground water in Florida and
the  experimental results of Lemley and Zhong (1983), it can be
assumed  that  the impact of ionic strength on  the  hydrolysis
rate  in  the citrus area is negligible.  Future studies  under
other  environmental conditions may show that ionic strength is
a more important factor.

The  effect  of  the  aquifer  material  on  surface  catalyzed
hydrolysis  rate  is  extremely  difficult  to  quantify.   The
variability  of  the  porous material in the  citrus  areas  of
Florida  is very great.  There is little information describing
the  influence  of specific surface types on hydrolysis  rates.
Some  studies  have  examined  hydrolysis  in  soil  and  water
systems  in  the  laboratory.  Table 3.7, from  Porter  et  al.
(1984),  compares  rates in the soil system to rates  estimated
using  distilled  water  studies.  The rates in soil  are  3-80
times  faster  than the rates calculated for  distilled  water.
Several  mechanisms have been theorized to explain this effect.
One  theory  suggests  that with the large surface  area  in  a
porous  medium  there  are  more surface  sites  available  for
reactions  to  take place.  Porter et al. (1984)  suggest  that
the  increase  in  the  hydrolysis rate could  be  due  to  the
buffering  capacity of the soil or catalysis by colloidal  soil
constituents.   Higher  microbial activity in soils could  also
account  for  the  higher rates.  However,  studies  done  with
sterile  and non-sterile samples in Long Island, New York  show
no  significant  differences in the degradation  rates   (Lemley
and  Zhong,  1984).  Although site specific studies  have  been
done  with  soil and ground water, it is difficult to  quantify
the  results  for  general  application to  the  estimation  of

                             143

-------
TABLE  3.7   COMPARISON OF DEGRADATION RATES OF ALDICARB RESIDUES ESTIMATED  BY
              HYDROLYSIS WITH  DEGRADATION  RATES MEASURED IN  SOIL AND WATER
              DEGRADATION  STUDIES   (Source:   Porter et  al.,  1984)
Sanple location
Long Island*

Wisconsin
Oerttal Sands
Resides
Presert
Suiroxlde
Suirore

Suiroxlce
Suirore
fiiproxlmate pH
or Sanple
an) 6

and 7
Suirore Only 7
Mirth Carol! ra
Coastal Plain
Suiroxlde
Suirore
and 5
Suirore Only 5
Florida Locatlora
Lull
Lake Hamilton
Alcana .
Ovletfj
Fort Pierce

Suiroxlce
Suirore
Suiroxlee
SuKore
Suiroxitfc
Suirore
SuUoxlde
Suirore

and 6
and 6
and 6
and 7
Suirore orly 7
Stud/
Tenperature
13
2S
25
25
25
25

25
25
25
25
25
Measured Rate
Cbf start,
(days'1)1
U.00076
0.0086
0.049
0.055
0.010
0.0047

0.014
0.015
0.014
0.25
0.024
(.00051)*
(0.0020)*
(0.0020)
(0.0042)
(O.U03)*
(0.0019)*

(0.0012)
(0.00079)
(0.0017)
(O.OJ9)
(0.0037)*
Measured Hair. Estimated Hair-Ure
Ufe Based on Distilled Estimated'
(ays)2 Nater H/crolysls (dtys) Measured
916
81
14
13
67
149

49
47
49
3
29 (
(351,
(
(
(
(
(

(
(
(
(
«,
12,
10,
62,
83,

»1,
40,
38,
1
a,
	 )• 3000 - 5000
177)* 600 - 1000
16) 80-160
16) 80
74)* 9000 - 6000
655)* 6000

61) 600 - 1000
55) 600 - 1000
63) 600 - 1000
,5) 80 - 160
47)* 80
4
10
y
6
80
40

16
17
16
40
3
  1. Hjmber InparerthEsls Is the starafird Aviation or data.

  2. Hjmbers Inparerthesls are the 99K corflderce Units.

  3. H>n-autoclaved sanples.

    •Prellmlrary estimate which "111 be rerired as more data becomes available.

-------
hydrolysis rates.

The  influence  of ionic strength and the aquifer  material  on
the  rate of hydrolysis of aldicarb is complex and still poorly
understood.   The  studies  that have been done show  that,  in
general,  these  two  factors  have the  competing  effects  on
hydrolysis.   The  presence  of competing  ions  depresses  the
rate, whereas the presence of colloids increases the rate.

The  effects of temperature and pH on the hydrolysis rate  have
been  extensively studied, such as those studies of Hanoen  and
Speigel  (1983)  and  Lemley and Zhong (1983),  sited  earlier.
Recently,  advances  have been made in understanding how  these
factors  influence the hydrolysis rate theoretically as well as
empirically.   In  general,  as  both the  temperature  and  pH
increase,   the   hydrolysis  rate  increases.   At   high   pH
concentrations,   the   number  of  OH   ions   available   for
hydrolysis  is greater, thereby increasing the likelihood of  a
reaction.   At  high  temperatures,  molecular  collisions  are
increased, therefore increasing the probability of a reaction.

As  in  the  unsaturated  zone, a  method  for  estimating  the
degradation  rate  based  on the environmental factors  in  the
Floridan   ground   water  is  necessary  to   estimate   model
parameters.   The same approach of a regression analysis  could
be  used  on  the existing temperature and pH data,  to  obtain
values  of  hydrolysis rate constants.  The analysis of  Lemley
and  Zhong  (1983)  suggests another approach.   They  use  the
Arrhenius   equation   to   describe  a   theoretically   based
relationship   between  the  second  order  reaction  rate  and
temperature.   In  this study, the Arrhenius  relationship  was
used  to  derive  the following equation  for  the  first-order
reaction  rate based on pH and temperature (see Appendix C  for
explanation and derivation):
                              145

-------
                     ~EA  X
          log k.  =	+ log A + pH - 14
               *   2.303R T
where     k  = first order reaction rate (/time)
          E a = activation energy (energy /mole)
                                                   o
          R  = universal gas constant (energy /mole  K)
          A  = pre-exponential factor (mole/liter-time)
          T  = temperature ( K)

The  activation energy (E ) and the pre-exponential factor  (A)
                         A
were  determined for aldicarb, sulfoxide and sulfone using  the
data  shown in Table 3.8.  This data was taken from studies  of
Hansen  and Speigel (1983), Lemley and Zhong (1983), Lemley and
Zhong  (1984), Porter et al .  (1984) and Lemley (1984,  personal
communication) .   Where  the second -order reaction  rates  were
not  reported directly (Hansen and Speigel, 1983 and Porter  et
al.,  1984), they were calculated from the data provided in the
studies  (Table  3.2  and  Table 3.9).  Table  3.10  shows  the
values  of  the  activation energy and ^pre-exponential  factor
resulting  from the data analysis.  Although fairly significant
errors  can result in these types of experimental analyses (see
Appendix  C),  there is good experimental agreement from  these
                                                              2
five  different  sources  of data as indicated by the  high  R
values (Table 3.10).
Temperature and pH Characteristics in Florida
The  temperature  in  the Floridan and  surficial  aquifers  is
fairly  constant, varying from 22 deg. C - 27 deg. C   (Stewart,
1966  and  Grain et al . , 1975) and averaging 25 deg. C   (Fayard
et  al.,  1983).   The pH is more variable.  In  the  surficial
aquifers,  pH varies from 5.0 to 8.0.  In the Floridan   Aquifer
the  pH  is rarely below 6.5 and sometimes is as high  as  9.0.

                             146

-------
 TABLE 3.8  HYDROLYSIS RATE DATA USED TO CALCULATE ACTIVATION
            ENERGY PARAMETERS
               Source
Aldicarb
Aldicarb
Aldicarb
Aldicarb
Aldicarb
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfoxide
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
Sulfone
A
E
A
B
C
A
E
A
B
C
D
A
D
D
D
A
B
C
B
A
B
C
D
A
B
C
D
D
D
B
B
 Second Order Rate
Constant, kr (liters/
      mole-min)

       .1637*
       .435
       .8946*
       .94
       1.15

       2.38*
       4.19
       13.77*
       10.5
       11.4
       4.38*
       23.5
       21.9*
       167*
       1110*

       4.35*
       13.6
       11.3
       19.1
       29.41*
       38.6
       33.0
       10.7*
       94.3*
       91.0
       90.5
       53.5*
       482
       2830*
       145
       192
                                                   Temperature
                                                        5
                                                        5
                                                       15
                                                       15
                                                       15

                                                        5
                                                        5
                                                       15
                                                       15
                                                       15
                                                       15
                                                       25
                                                       25
                                                       35
                                                       45

                                                        5
                                                        5
                                                        5
                                                       10
                                                       15
                                                       15
                                                       15
                                                       15
                                                       25
                                                       25
                                                       25
                                                       25
                                                       35
                                                       45
                                                       30
                                                       35
*Second order rate constant (kr) calculated by linear
 regression from data presented in given source.
A = Hansen and Spiegel, 1983
B = Lemley and Zhong, 1983
C = Lemley and Zhong, 1984
D = Porter et al., 1984
E = Lemley, personal communication, 1984
                             147

-------
           TABLE  3.9   DISTILLED WATER HYDROLYSIS  RATES  OF  ALDICARB  SULFOXIDE  AND ALDICARB
                         SULFONE  (Source:   Porter et al.,  1984)
*»
00
    Tenperature
      ( C)    4
AldLcarb Sulfoxide
      45    35 ( 34- 36)*
      35    74 ( 70- 78)•
      25   219 (207-232)*
      15    NA

Alcftcarb Sulfore
      45   109 ( 99-121)•
      35   396 (369-426)*
      25    NA
      15    NA
                                                                              Half-Life (days) at IndicatedpH
                                                                            7                 8
84 ( 74- 97) •
635 (527-798)*
NA
NA
106 ( 95-12U)*
584 (384-1218)*
NA
NA
31 ( 28- 34)*
130 (122-139)*
NA
NA
15 ( 14- 16)*
64 ( 62 - 67)*
NA
NA
4.4 (4.1-4.5)
2U ( 18- 22) •
16i (151-182)*
NA
1.7 (1.6-1.8)
8.2 (7.9-8.4)
82 ( 80- 83)*
420 (359-506)*
IM
2.9 (2.7-3.0)
25 ( 24- 27)
109 ( 99-122)*
N4
1.0 (0.9-1.1)
9.6 (9.4-y.B)
47 ( 45- 49)
N4
R
2.2
11 (
IM
R
0.9
4.5


(2.1-2.3)
10- 12)


(0.8-1.0)
(4.3-4.6)
              Numbers inparertheses represert the 93K Corflderce Interval
              N4 - rot measured
              NA - eroug^ data has  not yet been collected to estimate the hydrolysis rate
               R - half-life Is less than one day
               • - preliminary value, will be refined as more data become available

-------
    TABLE 3.10  ESTIMATED VALUES OF ACTIVATION ENERGIES AND
                PRE-EXPONENTIAL FACTORS FOR ALDICARB, ITS
                SULFOXIDE AND SULFONE
Pesticide
R
Activation
Energy, E,
(Kcal/mole)
Pre-exponential
   Factor, A
    (1/min)
Aldicarb      0.81

Aldicarb      0.92
  Sulfoxide

Aldicarb      0.90
  Sulfone
    20.8

    24.0


    21.8
                            6.31 x 10
                            1.58 x 10
                             1.0 x 10
                                                     15
                                     19
                                      18
                             149

-------
The  overall average for both aquifers based on the 1982  water
data  (Fayard et al., 1983) is 7.2.  The Floridan Aquifer alone
is  somewhat higher, the surficial aquifer is lower.  Based  on
the  average  pH and temperature and the activation energy  and
pre-exponential   factor   estimated  for  aldicarb   and   its
metabolites,  the  first  order reaction  rate  was  calculated
according to equation 3.5 (see Table 3.11).

In  addition to the four factors described above, Smelt et  al.
(1983)  have recently reported much higher disappearance  rates
for  aldicarb  sulfoxide  and aldicarb sulfone  when  incubated
under  anaerobic versus aerobic conditions. They reported  that
under   anaerobic   conditions,  disappearance  of  these   two
compounds  was from 8 to 100 times faster than in the same soil
under  aerobic  conditions.   More recent  work  (Smelt,  1984,
personal    communication   and   Bromilow,   1984,    personal
communication)   has  shown  that  in  anaerobic  soils  (redox
potentials  from  -60  mV  to +200  mV)  with  relatively  high
                                    O_L
concentrations  of ferrous iron (Fe  ), aldicarb is transformed
into  nitriles  and aldehydes.  Half lives for  aldicarb  under
these  conditions  were  0.5 days or less. Half lives  for  the
sulfoxide  and  sulfone in anaerobic, iron rich subsoils at  10
deg.  C ranged from 2 to 131 days.  The current theory is  that
the presence of Fe ^+ in solution catalyzes these reactions.

Very  little  specific data is available on redox potential  or
                             f\ i
dissolved  ferrous  iron (Fe   ) in Florida.  According to  Hem
(1970),  the  most common form of iron in the ground  water  is
ferrous  iron  (Fe *+) .  He states that ground water with  a  pH
beteen  6 and 8 can carry as much as 50 mg/1 of ferrous iron at
equilibrium  and  that the occurrence of 1.0 - 10 mg/1 of  iron
in  ground water is common.  Dissolved iron in the ground water
in  Florida varies from as little as 0.01 mg/1 to as high as 20
mg/1  (Fayard et al., 1983; Mycyk et al., 1983; U.S. Geological
Survey,  1982).   In general the surficial aquifer  has  higher

                             150

-------
 TABLE 3.11  FIRST-ORDER DECAY RATES IN FLORIDA'S GROUND WATER
             BASED ON ACTIVATION ENERGY ANALYSIS
Chemical       pH     °C     k. (days' )      Half Life  (days)
Aldicarb       7,. 2    25     7.91 x 10             876

Aldicarb
  sulfoxide    7.2    25     9.64 x 10             71.9

Aldicarb       7.2    25     2.34 x 10             29.7
  sulfone
                              151

-------
iron  concentrations because the aquifer materials contain more
iron-bearing   minerals   than  the   predominantly   limestone
Floridan Aquifer (Irwin, 1984, personal communication).

If  the  transformation of aldicarb described by Smelt  et  al.
(1983)  is indeed catalyzed by ferrous iron, the process  would
be  important  in the ground water in Florida.  At  this  time,
the  process is so poorly understood it can not be included  in
the  quantitative  analysis  of the  pesticide  parameters  for
modeling.
                              152

-------
                           SECTION 4
                 MODEL APPLICATION AND RESULTS
In  this  section  the application of the PRZM  and  the  CFEST
models  to  the unsaturated and saturated zones,  respectively,
is  discussed.  This discussion is broken down readily into two
separate  parts  because  the modeling of the  unsaturated  and
saturated  zones  was essentially  accomplished  independently.
The  unsaturated  zone model (PRZM) was applied under  each  of
the   scenarios  given  in  Section  2.1.6.   Daily   pesticide
loadings  at the bottom of the simulated soil core were written
to  an  output  file.  These loadings were summarized  and  are
presented  in this section.  Concurrently, steady-flow/unsteady
contaminant  transport  simulations were made using  the  CFEST
model  for  the saturated zone.  These simulations used a  unit
loading  of pesticide (1 kg/ha).  Because of this, results  are
expressed  as  relative concentrations in wells, that  is,  the
concentration  simulated  divided by the initial  concentration
in  the aquifer resulting from the input of the unit load.   In
the  last part of this section, results of the unsaturated  and
saturated   zone  modeling  are  combined  to  predict   actual
drinking well-water concentrations.
                              153

-------
4.1  UNSATURATED ZONE MODELING
Before   discussing   the  results  of  the  unsaturated   zone
modeling,   the  process  of  applying  the  model,   including
assumptions  made,  will be covered.  Part of this process  was
to  attempt  to  verify  the  PRZM  model  using  residue  data
available for Florida citrus soils.
4.1.1  PRZM Verification
Two  data  sets were available for verification purposes.   One
has  been  reported in Jones et al. (1983) and Hornsby  et  al.
(1983).   These  data  were  taken from two  sites  in  central
Florida;  one a typical "ridge" citrus site  (Lake Hamilton) and
the  other  a typical "flatwoods" site (Oviedo).  Another  data
set  was  also  produced  as  a  part  of  this  study  by  the
University of Florida (Foran, 1984; personal communication).

4.1.1.1  1983 Residue Data—

The  first  data  set  was  used in  the  following  way.   The
currently  available  version of PRZM  (PRZM-II, Carsel et  al.,
1984)  was  run  using data inputs which were  exactly   (or   as
close  as  possible, considering input changes) those  used   by
Jones  et  al.  (1983) in their original PRZM model  calibration
exercise.   This was done because Jones et al. used an  earlier
version  of PRZM in their work.  Model outputs were compared  to
outputs  produced  by  Jones  and measured field  data  at  the
Oviedo  and Lake Hamilton sites.  This done, another simulation
was  performed.   In this case PRZM-II, as modified for use   in
this  study,  with  parent/daughter relationships  and   lateral
drainage  was used.  Inputs used were  those  of the generic soil
                              154

-------
orders  as  generated by the analysis described in  Section  2.
At  the  Lake  Hamilton site, the dominant soil  order  is  the
entisols.   Therefore,  input  data to be used  for  simulating
entisols   in  the  unsaturated  zone  scenarios  were   input.
Likewise,  the input sequence to be used for spodosols was used
at  the  Oviedo site.  Once again, results are compared to  the
observed  data  and model results from Hornsby et  al.  (1983),
and  Jones  et al. (1983).  Results for the Lake Hamilton  site
are shown in Figures 4.1 through 4.3.

Figure  4.1 shows a comparison of the amount of aldicarb (total
toxic  residue)  remaining in the top 300 cm of the profile  at
Lake  Hamilton  up  to 200 days after  application.   Both  the
model  calibration  of  Jones  and the  model  simulation  with
PRZM-II  agree  well  with the observed  data.   PRZM-II  shows
consistently  slightly  less residues than the earlier  version
of  PRZM  used by Jones.  The generic input parameter  set  for
entisols  shows  more  rapid  decay of  the  TTR   (total  toxic
residue)  than observed in the field or predicted by the  other
modeling  exercises.   Figure  4.2 shows a  comparison  of  the
depth  of  the  simulated  peak from  the  models  and  deepest
penetration  observed  in  the  data.  The  PRZM-II  code  with
Jones'  inputs and the PRZM-II code with generic inputs  under-
simulate   both  the  data  and  the  original  PRZM  up  until
approximately  40  days, at which time they follow the  deepest
penetration   observed  in  the  data.   The  reason  for   the
discrepancy  between Jones' PRZM simulation and the simulations
in  this study soon after pesticide application appears to be a
percolation  event  which  occurred shortly  thereafter.   This
event  caused movement of the chemical in Jones' simulation but
not  in  ours.  In both cases, Lake Alfred  Experiment  Station
1983  precipitation data was used.  The discrepancy may not  be
in  the  precipitation record but in initial  conditions.   Our
simulations  began January 1 so initial conditions were set  by
the  model.   These  particulars of Jones' simulation  are  not

                             155

-------
   100
    80-
UJ
ec
o
o
o


2

<


Ul
cc

Ul
o
m
la
a.
                    50            100



                          DAYS AFTER APPLICATION
Figure  4.1
Predicted  and observed Aldicarb  TTR in the

upper  300  cm of the soil at  the  Lake Hamilton

site,  1984.   (After Jones et al.,  1983).
                             156

-------
   300-

                                     0
                                                       120
                       DAYS AFTER APPLICATION
                           LEGEND

                        PRZM

                      - MODIFIED PRZM - II

                        PRZM - II

                        DEEPEST PENETRATION MEASURED
Figure  4.2  Predicted movement of  Aldicarb residues at
             the Lake Hamilton location (After  Jones
             et al.,  1983).
                           157

-------
known.   Figure 4.3 shows a comparison of the depths over which
aldicarb  TTR  >5  g/g (parts per billion) was observed in  the
field  versus  our  two model simulations.  We note  here  that
aldicarb  was  applied on February 16, 1983.  The  four  shaded
areas  at  the  left  of each date are samples  taken  in  four
quadrants  of  the  field.   At the March 4  date,  both  model
simulations  agree well with the observed data.  Note that with
the  PRZM-II  model,  as modified for this study,  the  TTR  is
broken  down  into  aldicarb, aldicarb sulfoxide  and  aldicarb
sulfone.   At the April 6 date, both models and the data  still
agree  well.   Notice  that no aldicarb is simulated  as  being
present  at this date.  On May 3, PRZM-II still agrees well but
the  modified PRZM-II agrees less well.  This may be due either
to  too  rapid movement or too rapid decay.  The same  argument
holds for the June 15 sampling date.

Figures  4.4 and 4.5 show the results of model applications  at
the  Oviedo  (flatwoods)  site.  Figure 4.4  compares  the  TTR
remaining  in the top 150 cm of the soil profile.  The measured
decay  in this case does not appear to be first-order.   Jones'
simulation  and  the modified PRZM-II simulation  with  generic
spodosol  inputs  give  very close to the  same  results.   The
PRZM-II  model  with  Jones' inputs oversimulate  the  quantity
remaining in the profile.

Figure  4.5  shows the comparison of observed TTR in the  field
versus  PRZM-II  with Jones' inputs and modified  PRZM-II  with
generic  spodosol  inputs.  Agreement between the  position  of
residues  in  the  field and both model simulations  are  good.
The   modified  PRZM-II  model  with  generic  spodosol  inputs
appears   to  be  slightly  better.   This  is  due  to  better
agreement  of the decay rates in this model with those observed
in  the  field.   Notice  that  only sulfone  is  left  in  the
simulated profile on the June 14 sampling date.
                              158

-------
          3/4/83
4/6/83
5/3/83
                                                           6/15/83
  100-
o.
UJ
Q
  200-
  300 J
                                 OBSERVED


                                 PRZM-II


                                 MODIFIED PRZM-II


                                 NO SIMULATED RESIDUE > 5ng/g
 Figure  4.3  Location of Aldicarb TTR  in the  soil profile

              > 5  ppb at  the Lake  Hamilton  site.
                                  159

-------
     100
 UJ
 cr
 o
 u
 o
 z
  5
  UJ
  oc

  UJ
  o
  CO
  UJ
  Q.
                 OBSERVED
     20-
                                   100


                            DAYS AFTER APPLICATION
Figure  4.4
Predicted  and observed Aldicarb TTR in  the  upper

150 cm of  the soil at the  Oviedo site.   (After

Jones et al., 1983).
                              160

-------
       3/3/83
4/5/83
6/2/83
6/14/83
150-
                                    OBSERVED
                             f'.--•-••.1  PRZM- II

                                    MODIFIED PRZM-II
 Figure 4.5   Location of Aldicarb  TTR in the  soil profile
               > 5 ppb at Oviedo.
                                   161

-------
These  results indicate a fair agreement between the PRZM model
and  the  observed  data  from the 1983  sampling  of  aldicarb
residues  at  Lake Hamilton and a good agreement at the  Oviedo
site  (Hornsby et al., 1983; Jones et al., 1983).   Simulations
with    PRZM-II    modified   to    simulate    parent/daughter
relationships  were  also fair to good.  The  generic  entisols
input  data  set produced higher than observed  degradation  at
the  Lake Hamilton site, while the generic spodosol input  data
produced  very good agreement with observed residue data at the
Oviedo  site.  The agreement in both cases is suprisingly  good
considering  that site specific soils and pesticide  parameters
were  not  used  at either site.  This somewhat  justifies  the
assumption   that  the  soil  characterization  and   pesticide
degradation  and transformation rate data analysis can be  used
to  represent  aldicarb  fate and transport  processes  in  the
citrus growing region of Florida.

4.1.1.2  1984 Residue Data—

Other  data  were also available for 1984 from the  Oviedo  and
Davenport  sites.   The  Davenport  site is  located  near  the
intersection  of  Interstate Highways 4 and 75 and  is,  again,
typical  of  a "ridge" citrus site.  The residue  samples  were
collected  by  personnel  at the University  of  Florida  under
contract  to  the  U.S. Environmental Protection  Agency.   The
samples  were  analyzed  by the U.S.  Environmental  Protection
Agency  in  Beltsville, Maryland.  The data, as  received,  are
shown  in  Appendix  D.  An analysis of the Davenport  data  is
shown in Table 4.1.

This  data  has  two  interesting features.   First,  the  peak
concentration  on  5/9/84  is deeper than the peak  sampled  on
6/13/84.   Although  this peak is shown on 5/9/84 in the  365.8
to  426.7  cm  range, and on 6/13/84 in the 304.8 to  365.8  cm
range,  the  peaks  may  actually be  quite  close.   The  more

                             162

-------
 TABLE 4.1  ALDICARB RESIDUES (TTR) IN THE SOIL AT DAVENPORT,
            FLORIDA, 1984
                                    Mean TTR (ppb)*
 Depth (cm)                   5/9/84                 6/13/84

    0-30.5                    47.33                   10.50

 30.5-61.0                     0.00                    5.00

 61.0-121.9                    6.00                   22.75

121.9-182.9                   12.67                   48.50

182.9-243.8                   42.00                   88.50

243.8-304.8                   78.67                   91.00

304.8-365.8                   70.67                  108.00

365.8-426.7                  106.00                   92.50

426.7-487.7                   47.33                   53.00
Mean TTR in profile           48.37                   64.00
*Values used to compute the means were only those determined
 to be within the treated band.
                             163

-------
troubling  aspect  is  that  the  peaks  are  almost  the  same
magnitude  although  more than a month had elapsed between  the
two  sampling dates.  In fact, the data shows greater than  25%
more  aldicarb  TTR in the profile on 6/13/84 than  on  5/9/84.
These differences are not easily explained.

The  data  from  the  1984 Oviedo sampling  are  very  erratic.
There  are many non-detectable entries in the data so that  the
mean  of  the  data  from various soil cores  may  be  somewhat
misleading.   However,  the mean values are presented in  Table
4.2 for comparison with the Davenport data.

The  TTR  in  the soil at the 5/22/84 sampling data  at  Oviedo ,
compares  favorably  with  that in the soil and  the  Davenport
site  on 5/9/84.  The average TTR in the soil at Oviedo on  the
6/3/84  sampling date of 18.4 ppb also compares favorably  with
the  mean  TTR concentration at Davenport in the top 152 cm  of
soil  on 6/13/84 of 21.7 ppb.  However, considering the erratic
nature  of the 1984 data from Oviedo, this favorable comparison
may  be  fortuitous.  Indeed, the pesticide was applied a  full
month  earlier at Davenport  (2/14/84) than at Oviedo (3/14/84).
Therefore,  to  have roughly the same quantity of pesticide  in
the  soil on the sampling dates, the degradation rates have  to
be higher at Oviedo.

An  analysis of the degradation rates are shown in Figure  4.6.
The   dashed  lines  on  the  plot  correspond  to  half-lives,
reported  by  Hornsby et al.  (1983), of 3 to 46 days at  Oviedo
and  Lake  Hamilton, Florida.  Assuming that  little  pesticide
leached   past  the  150  cm  depth  at  Oviedo  in  1984,  the
approximate  percent  TTR  remaining  in  the  profile  can  be
determined  from  the  data  in Table 4.2 for the  two  sampling
dates.    (A  bulk  density  of  1.5  g/cm   was  used  in  this
calculation).   The logarithms of these percentages are plotted
on  Figure  4.6  along with  the best fit line.   This  gives   a

                              164

-------
TABLE 4.2  ALDICARB RESIDUES (TTR) IN THE SOIL AT OVIEDO,
           FLORIDA, 1984
                                      Mean TTR (ppb)
 Depth (cm)                  5/22/84                  6/20/84

    0-30-5                    173.5                     20.4

 30.5-61.0                     29.8                      2.7

 61.0-91.4                      8.0                     42.2

 91.4-121.9                    23.0                     21.3

121.9-152.4                     0.0                      5.3
Mean TTR in profile            46.9                     18.4
                              165

-------
              10  20  30  40 SO  60  70  80  90 100 110 120
                         DAYS AFTER APPLICATION

                   O  OBSERVED AT DAVENPORT. 1884

                   X  OBSERVED AT OVIEDO. 1884

Figure  4.6  Comparison of Aldicarb TTR degradation rates for
             1984 Oviedo and Davenport data.
                               166

-------
half-life   of   approximately  28  days  for   aldicarb   TTR.
Similarly,  the data from Davenport are plotted.  The half-life
of  these residues is 210 days which is five times greater than
the  fastest  rate from the Lake Hamilton/Oviedo  1983  studies
and  roughly  10 times greater than rates determined  from  the
1984  Oviedo data.  If, in fact, these determined rates for the
soils  at  Davenport  are valid, they are at best  three  times
slower  than  reported values of Hornsby et al. (1983), or  Rao
et al. (1984), for several Georgia coastal plains soils.

The  modified PRZM-II model was applied to these two data  sets
in   a   verification   exercise.    For   the   Oviedo   site,
precipitation  data  from  Sanford experiment station  and  pan
evaporation  from  Lisbon  were used  as  model  meteorological
inputs.   The  generic  spodosol input parameter  sequence  was
used  for soils, crop and pesticide parameter inputs.   Results
are  shown in Figure 4.7.  The mean values of the measured data
are  shown  as  vertical bars.  The dashed lines indicate  a  1
standard  deviation range calculated from the data.  The  model
simulations  follow  the  mean values fairly  closely  and  are
certainly  within  a 1 standard deviation distance  around  the
means  at each depth.  The model under-simulated the total mass
in  the profile by about 40% on the May 22 date and 45% on June
20.   Therefore,  model degradation rates seem somewhat  higher
than those observed.

The  modified  PRZM-II model was also applied to the  Davenport
site.   February  through  June, 1984, Lake  Alfred  experiment
station  precipitation  and pan evaporation data were  used  as
meteorological  inputs  and  the generic entisol data  set  was
used  for  soils,  crop  and pesticide  parameter  input.   The
results,  shown in Figure 4.8 indicate a radical difference  in
simulated  and observed pesticide concentrations.  This is  not
unexpected   since  the  pesticide  degradation  rates  in  the
generic   entisol   parameter  data  set  are  more   than   an

                              167

-------
CO
                               CONCETRATION (ppb)
                              200      300      400
                                         MAY 22. 1984
                               TI11 J  MEASURED
                                0——O   SIMULATED
500
600   0
CONCENTRATION (ppb)
    100      200
                       JUNE 20. 1964
                                                                       ,.!,.	1
                                                                       _ 	J
        Figure  4.7  Comparison of measured and simulated Aldicarb TTR concentrations
                     for two 1984 sampling dates  at Oviedo, Florida.

-------
                   CONCENTRATION (ppb)


                   100     200     300
                                               CONCENTRATION (ppb)

60


100

1-lJ
£
H"^

1 1
[ | MAY 9. 1984
J"
160 ii ;
hi-'i--
Ai 1 i
200?
      ^  250

      X



         300
VO
350





400 6 '




450




600
  •-,-fr-'

   1     ;
   ,1



IQT
                                      MEASURED


                                      SIMULATED
                                       0
                                                         200
                                                          i
300
                                                                JUNE 13. 1984
       Figure 4.8
        Comparison  of  observed  and  simulated Aldicarb TTR concentrations

        for  two  1984 sampling dates at  Davenport, Florida.

-------
order-of-magnitude  greater  than the observed TTR  degradation
rates.   Another perplexing aspect of this data has to do  with
the  depth of penetration of the peak.  The data shows the peak
on  9 May 1984, at about the 60-420 cm depth in the soil.  Even
when  the  PRZM-II  degradation rates were  adjusted  to  match
those  observed  in  the field, the simulated position  of  the
peak  on  9 May was approximately 30 cm.  On the 13 June  date,
the  simulated  peak  was  between  100  and  160  cm.   It  is
possible,  however,  that much more rainfall fell at this  site
than is indicated by the Lake Alfred data.

4.1.1.3  Verification Summary—

Verification  of the PRZM-II model was performed using  residue
data  from four sampling dates at three different locations  in
Florida.   Generic  soils, crop and pesticide  parameter  input
data  were used as opposed to specific data for the  individual
sites.   Thus, the intent was to verify not only the model, but
also  the  generalized data sets to be used for  simulation  of
aldicarb  fate  and transport in the various  unsaturated  zone
scenarios.

Verification   results  were  best  at  the  Oviedo  site  were
spodosolic  soils  are present.  The model simulations  matched
observed  concentration  profiles well for both 1983 and  1984.
At  the  Lake Hamilton and Davenport sites, when  "ridge"  type
soils  occur,  simulation results were not as good.   At  these
locations,  too  rapid degradation of the pesticide appears  to
be  simulated.  Results at the Lake Hamilton site were  better,
however,  than  those at the Davenport site.  Some  aspects  of
the  1984 Davenport data; the extremely low decay rates and the
deep  penetration of the peak concentration were not able to be
explained.
                             170

-------
4.1.2  Model Application and Assumptions
The  verified  PRZM-II  model  was  applied  for  each  of  the
twenty-four  scenarios  listed in Section 2.  Simulations  with
overhead   irrigation   were  made  first.   The  mean   annual
evapotranspiration  was calibrated to approximately 122 cm  (48
in.)   by   adjusting  the  available  water  level  at   which
irrigations  were  triggered.  For each soil order,  this  came
out  to  be roughly 45% (i.e., 45% of the water  between  field
capacity  and wilting point).  Other than this, no  adjustments
were  made to any parameter in the generic input sequences  for
the no-irrigation or overhead irrigation practices.

4.1.2.1  Simulation of Low Volume Spray Irrigation—

For  the simulation of the low volume spray irrigation practice
application  of  the  model was not  as  straightforward.   The
problem  lies in the fact that PRZM is a one-dimensional  model
and  the simulation of the low volume spray (or any  irrigation
practice   in  which  the  irrigation  water  is  not   applied
approximately  uniformly over the simulated area) is a  spatial
problem.   A schematic of a unit "ridge" citrus block is  shown
in  Figure  4.9.   This block contains one tree.   Aldicarb  is
applied  centered  beneath the drip line on either side of  the
canopy.   Depending on the size of the tree and location of the
aldicarb  band, the wetted area of the spray jet may or may not
intersect  the treated band.  If it does, there are two  unique
areas with respect to aldicarb leaching;

    1)   the area in which only rainfall falls upon the treated
         band, and

    2)   the area which receives rainfall plus irrigation
         water.

                             171

-------
                         5.2 m
Figure 4.9  Schematic  of  a  unit block of citrus (one
            tree,  not  to  scale).
                          172

-------
Notice  that  if spray jets are located down the line of  trees
so  that  their wetted patterns overlap, then a wetted band  as
opposed to a circle develops.

Considering  these two areas together, the leached from a  unit
citrus block would be:

           L = W-L L! + W2 L2                             (4.1)

where     L, = load from the non-irrigated portion of the
               treated band
                                                      band
     L2 = load from the irrigated portion of the

W, ,  W- = fractions of the unit block made up of
          areas 1 and 2
It  was  felt  that PRZM-II could be applied to each  of  these
areas and the results weighted to obtain the proper load.

In  order to properly simulate pesticide movement in each area,
the  hydrology  of each area must be properly simulated.   This
means  that  evapotranspiration  from areas which are  irrigated
and  not  irrigated  should  be  adjusted  so  that  the  total
evapotranspiration  will equal that required on an annual basis
(  122 cm).  Therefore:

            ET = Vli ET-L + W2 ET2                          (4.2)

The  problem  here is that the weights, W, and W_ are  unknown.
A  best  guess,  however, is that they correspond to  the  root
distribution  in each area.  This too is unknown.  For the sake
of  simplicity, the assumption was made that all the roots were
in  the wetted zone  (i.e., W_ =  1.0).  This means that all   the

                              173

-------
evapotranspiration   demand   for  the  entire  tree  must   be
extracted  from  the  wetted band.   Under  uniform  irrigation
                                                              2
water  application, the demand is 122 cm over the 7.6 x 7.6  m
unit  block.   Therefore, if all the demand must come from  the
wetted areas (10 m2) the demand must be increased by:
                     7.6 x 7. 6 m2                        (4.3)
                         10 m2

or  a factor of 5.8 in the ridge areas.  In the flatwoods areas
tree  spacing  of  4.6 m x 7.6 m was  assumed,  therefore,  the
demand  must  be increased by 3.51.  To accomplish this in  the
simulations,  the  input  pan  evaporation  factor  (PFAC)  was
multiplied by the above values as needed.

4.1.2.2  Simulation of Drainage in Bedded Citrus—

For  spodo.sols  and alfisols, the lateral drainage from  bedded
citrus  was simulated.  A discussion of the algorithms used  to
accomplish  this is discussed in Appendix A.   Parameterization
of the algorithms was discussed in Section 2.

The   water   draining  laterally  from  the  soil  may   carry
pesticide,  just  as the percolating water does.   To  simulate
the  removal  of  pesticide from the unit  citrus  block  would
require  that  a  sink  term be added  to  the  one-dimensional
advection/dispersion   equation   in   PRZM   for   each   soil
compartment, much like the runoff term for the surface layer.

Again,  this  problem  is  at least  two-dimensional.   As  the
pesticide  moves laterally out of the unit block, some will  be
adsorbed  by  soil.  This adsorbed chemical may be  subject  to
later desorption and leaching by rainfall.
                              174

-------
Even  pesticide which enters adjacent canals may be subject  to
leaching.   As  a "worst case" it was assumed that all  of  the
pesticide,  whether or not advected by lateral flow, still  has
the  potential to move to the ground water, until it  degrades.
For  this  reason, lateral flow was allowed to leave  the  unit
block  without  carrying  pesticide.   Using  this  assumption,
lateral  drainage  can  still be simulated for the  purpose  of
simulating  a  proper  water balance while providing  a  "worst
case" scenario for pesticide leaching.

4.1.2.3  Pesticide Parameter Sensitivity—

No  sensitivity  analyses were performed on  transformation  or
degradation  rates or adsorption partition coefficients for the
unsaturated  zone.   For the "flatwoods" type soils  (spodosols
and  alfisols) the verification results using the generic input
data  sets were good and sensitivity analysis was deemed to  be
unnecessary.   For  the entisols and ultisols, it appears  that
use  of slower degradation rates might be  in order.  This  will
be discussed further later on in this section.
4.1.3  Hydrologic/Hydraulic Results
Table  4.3 shows the water balance results for the  twenty-four
unsaturated zone scenarios.

Evapotranspiration   (from  the  canopy and  the  soil  profile)
averaged  81.1  cm  (31.9 in) over the non-irrigation  scenarios
for  all  soil  orders.  Evapotranspiration averaged  121.2  cm
(47.7  in) over the overhead irrigation scenarios for all  soil
orders.   As mentioned earlier, these depths were   "calibrated"
to  equal  about 122 cm  (48 in), which is a typical  value   for
irrigated  Florida  citrus.  The irrigation by low volume  spray
                              175

-------
TABLE 4.3  SOME WATER BALANCE COMPONENTS FOR TWENTY-FOUR
           UNSATURATED ZONE SCENARIOS (in centimeters).
           Values are annual means over 14 years
Scenario
EN2.7
EN9.0
E02.7
E09.0
EL2.7
EL9.0
UNI. 8
UN9.0
UO1.8
U09.0
UL1.8
UL9.0
SN1.2L
SO1.2L
SL1.2L
SN1.2H
S01.2H
SL1.2H
AN1 . 2L
A01.2L
AL1 . 2L
AN1 . 2H
A01 . 2H
AL1 . 2H
Evapotrans-
piration
82.5
82.8
124.4
124.5
518.3
525.8
86.7
86.6
125.6
125.6
567.2
570.9
80.0
124.6
355.5
79.2
113.2
333.6
75.9
121.6
343.2
75.0
110.3
322.0
Irrigation
__
—
60.3
59.7
436.7
444.2
—
—
58.6
58.7
480.6
486.1
—
61.7
259.6
—
45.7
227.0
—
61.8
240.6
—
48.5
212.7
Recharge
38.7
38.4
57.1
56.4
39.5
39.6
34.5
35.1
54.2
54.7
34.5
36.4
33.0
47.1
20.9
44.7
52.2
27.1
30.4
29.7
12.3
26.0
30.1
15.3
Lateral
Drainage

—
—
—
—
—
—
—
—
—
—
—
8.2
11.2
4.4
15.8
20.0
6.1
14.9
31.7
6.3
38.7
47.8
15.1
                           176

-------
yielded  evapotranspiration  depths  in the wetted  area  which
ranged  from 322 cm (127 in) to 570.9 cm (224 in).  On a  total
area  basis,  assuming  no evapotranspiration  from  (i.e.,  no
plant  roots  in) the non-irrigated areas,  this depth is 93  cm
(36.8  in)  to  97  cm  (38.2  in).   It  appears,  then,  that
evapotranspiration  depths  are on the order of 20 to  23%  too
low  for the low volume spray scenarios on a total area  basis.
A  possible  explanation  for this is that the method  used  to
simulate  transpiration  with  PRZM-II created  water  limiting
situations  in  the  soil profile, even with  the  addition  of
irrigation  water.   That  is, there is  not  enough  available
water  in the root zone at field capacity to satisify this high
demand.   Therefore,  not all the potential  evapotranspiration
was  used  each day.  In reality, because of the  abundance  of
rainfall,  roots that extract water from the soil by the  plant
probably  occur outside the wetted area.  If the assumption  is
made,  however,  that  the  plant roots  take  water  from  the
non-irrigated   area   as   well  at  rates  taken   from   the
non-irrigated    model    simulations,   then   the    weighted
evapotranspiration  depths  are too large, ranging from 150  cm
(59 in) to 169 cm (66 in).

Thus,  it  appears,  to accurately simulate the  hydrology  and
hence  the  movement  of pesticide under the low  volume  spray
scenario,  three areas would have to be considered rather  than
the two areas originally considered in Section 4.1.2.1;

    1)   the irrigated area including plant roots and
         accompanying water and pesticide uptake

    2)   the non-irrigated area having plant roots and
         accompanying water and pesticide uptake and

    3)   a non-irrigated area having no plant roots and no
         transpirational or pesticide uptake.
                              177

-------
The  weighted  average  evapotranspiration,  ETw,  for  a  unit
citrus block would be:

            ETW = WJL  ETX + W2 ET2 + W3 ET3               (4.4)

where  the subscripts 1, 2,  and 3 refer to the areas above  and
W-  are  the  weights  for each area.   A  rational  method  of
assigning  these weights would be to use the root  distribution
in  the  three areas.  However, these values are  unknown,  and
there is no simple way of determining these weights.

Because  of this difficulty, it was felt that interpretation of
results  based on one low volume spray simulation would not  be
meaningful,    therefore,   they   were   not   given   further
consideration.

Irrigation  using  overhead  methods  resulted  in  an  average
annual  water  application of 56.9 cm (22.4 in) over  all  soil
orders.   This  application  rate is higher  than  the  average
values  computed from the data of Duerr and Trommer (1982)  for
overhead  systems  in the  state, of roughly 38  cm  (15  in).
However,   they   do  record  data  yielding   annual   average
application  rates  as high as 25.4 in. per  acre   (1970-1980).
Therefore the simulated depths do not seem unreasonable.

Average  annual recharge rates under no irrigation ranged  from
a  low  of 12.3 cm/yr (4.8 in/yr) for low  rainfall alfisols  to
44.7  cm/yr (17.6 in/yr) for high rainfall spodosols.  The mean
for  no-irrigation  over all soil orders was 35.1  cm/yr   (13.8
in/yr).    These  rates  are  in  good  agreement  with   those
mentioned  in Section 2 derived from the literature.   Recharge
rates increased when  irrigation water was  applied.

Lateral  drainage,  of  course, only occurred in  alfisols  and

                              178

-------
spodosols.   Spodosols  exhibited the lowest  lateral  drainage
depths  with  average annual values being 7.9 cm (3.1  in)  for
low   rainfall,  and  14.0  cm  (5.5  in)  for  high   rainfall
spodosols.   This represents approximately 6.5% of the  average
annual  precipitation  over the 14 year simulation  period  for
the  low  rainfall  spodosols  and 10% for  the  high  rainfall
spodosols.   The  alfisols  had still higher  lateral  drainage
losses,  being  7.6 cm (6.9 in) for the low rainfall  scenarios
and  3.9  cm (13.3 in) for the high rainfall  scenarios.   This
represents  14%  and  24%  of the low  and  high  precipitation
depths, respectively.

Examination  of the lateral drainage and recharge rates reveals
that  a large percentage of the total water that exits the soil
profile  due to lateral drainage and recharge exits  laterally.
In  the spodosols, the percentages range from 17 to 28%,  while
in  the  alfisols the percentages range from 33 to  61%.   This
indicates  that  the  contamination of surface  water  is  also
quite likely in areas where these soils are prevalent.
4.1.4  Pesticide Fate and Transport Results
Figure  4.10  graphically depicts the overall fate of  aldicarb
TTR  after  application to the soil.  The values presented  are
annual  average  percentages  of applied chemical over  the  14
year  period.   Three  processes  are  represented;  uptake  by
plants,  degradation to non-toxic residues and leaching to  the
saturated  zone.   By  far, the most eminent  fate  process  is
decay  in  the root zone.  Over all scenarios,  this  accounted
for  an  average  of  70.3% of  the  applied  chemical.   Plant
uptake,  on  the  average, accounted for 29.2% of  the  applied
chemical.   The  remainder,  or 0.5% of the  applied  chemical,
leached  to  the  saturated zone on an annual  basis  over  all
                              179

-------
     PERCENT OF APPLICATION
0  10  20 SO 40  60  60  70  80 8.0 100
                                 NO IRRIGATION




                                 OVERHEAD
6 FOOT CORE
                                 NO IRRIGATION




                                 OVERHEAD
                                                           ULTISOLS
30 FOOT CORE
                                 NO IRRIGATION




                                 OVERHEAD
9 FOOT CORE
                                 NO IRRIGATION




                                 OVERHEAD
                                                          ENTISOLS
30 FOOT CORE
                                 NO IRRIGATION




                                 OVERHEAD
LOW RAINFALL
                                                           ALFISOLS
                                 NO IRRIGATION




                                 OVERHEAD
HIGH RAINFALL
                                 NO IRRIGATION




                                 OVERHEAD
LOW RAINFALL
                                                           SPODOSOLS
                                 NO IRRIGATION




                                 •OVERHEAD
HIGH RAINFALL
                                  PLANT UPTAKE





                                  DECAY TO NONTOXICS




                                  LEACHED TO GROUNDWATER
 Figure  4.10   Fate of Aldicarb TTR  after application to  soil,
                                  180

-------
scenarios.    Table   4.4  gives  the   simulated   percentages
accounted for by the three processes.

Figure  4.11  takes  a  more detailed  look  at  the  pesticide
leached  under  each  scenario.   Shown in the  figure  is  the
geometric  mean  quantity of aldicarb TTR leached to  saturated
zone  over the 14 year period, expressed in Kg/ha, for each  of
the  sixteen  unsaturated  zone scenarios.   The  highest  mean
annual  loss  occurs with overhead irrigation with a 270 cm  (9
ft)   unsaturated   zone  thickness  (.076   Kg/ha-yr).    This
represents  1.3%  of  the  applied  5.6  Kg/ha.   Next  is  the
ultisol,  overhead irrigation, with a 180 cm (6 ft) unsaturated
zone  thickness.  Undoubtedly, highest loads occur on  entisols
and  ultisols  with thin unsaturated zones, regardless  of  the
irrigation  method  used.   When thick  unsaturated  zones  are
present  in  combination  with these soils, the  loads  are  in
general  lower  than from the spodosols.  Lowest leached  loads
are associated with alfisols.

It  is  obvious  from  looking at Figure 4.11  that  there  are
differences  among the mean responses from various soil orders,
unsaturated   zone  depths,  irrigation  methods  and  climatic
scenarios.   The  effect  of  unsaturated  zone  depth  in  the
entisols  and ultisols and the effect of soil order seem to  be
the  most  pronounced.   Irrigation  method seems  to  be  less
important  as  does high or low rainfall in the  spodosols  and
alfisols.   Initially, these scenarios were delineated  because
it  was  felt  that  significantly  different  responses  would
result.   To test this assumption several analyses of  variance
(ANOVA) tests were performed.

All  ANOVA  tests performed were one way.  Tests were  done  on
all  "treatments,"  that  is, soil  order,  irrigation  method,
unsaturated  zone  thickness  (entisols and ultisols  only)  and
high  and  low rainfall  (spodosols and alfisols only).   As  an

                              181

-------
                     TABLE  4.4   FATE OF  ALDICARB TTR AFTER  APPLICATION TO SOIL
                                                                 PERCENT OF APPLICATION  (in block)
                   Soil Type


                   ULTISOLS
                   ENTISOLS
00
to
                   ALFISOLS
                   SPODOSOLS
                                   6 FOOT CORE
                                  30 FOOT CORE
                                   9 FOOT CORE
                                  30 FOOT CORE
                                  LOW RAINFALL
                                 HIGH RAINFALL
                                  LOW RAINFALL
                                 HIGH RAINFALL
  Irrigation
No
Overhead
Low Vol.

No
Overhead
Low Vol.
                                                           Spray
                                                           Spray
No
Overhead
Low Vol. Spray

No
Overhead
Low Vol. Spray
No
Overhead
Low Vol. Spray

No
Overhead
Low Vol. Spray
No
Overhead
Low Vol. Spray

No
Overhead
Low Vol. Spray
Plant
Uptake
15.57
21.70
54.97
6.91
13.57
45.46
14.36
22.44
56.51
9.38
17.61
51.17
15.33
25.28
45.31
12.55
24.74
48.31
18.13
30.09
51.35
15.54
28.76
54.46
Decay
83.70
76.95
44.17
93.08
86.40
54.53
83.80
74.49
41.54
90.55
82.26
48.79
84.66
74.70
51.68
87.40
75.21
51.68
81.80
69.82
48.64
84.19
70.97
45.49
Leached
0.73
1.35
0.86
0.01
0.03
0.01
1.84
3.07
1.95
0.07
0.13
0.04
0.01
0.03
0.01
0.05
0.05
0.01
0.07
0.09
0.01
0.27
0.27
0.05

-------
CO
to
                10-°
MEAN ANNUAL PESTCIOE LOAD LEACHED TO SATURATED ZONE (KG/HA)


 10'4         10-3         10-*         10-1
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    6 FOOT CORE
                                                                                             ULTISOLS
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    30 FOOT CORE
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    9 FOOT CORE
                                                                                             ENTISOLS

                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                   30 FOOT CORE
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    LOW RAINFALL
                                                                                             ALFISOLS
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    HIGH RAINFALL
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    LOW RAINFALL
                                                                                             SPODOSOLS
                                                                    NO IRRIGATION

                                                                    OVERHEAD
                                                    HIGH RAINFALL
                 Figure 4.11  Geometric mean  annual quantity of  pesticide  leached to the
                                saturated zone  from  the  treated band.

-------
example,  the  one way ANOVA of irrigation method is  shown  in
Figure  4.12.  The data are grouped by irrigation method.   The
ANOVA  table  shows that the computed F value of 0.91 does  not
exceed  the critical value of 3.1 at the 90% confidence levels.
Therefore,  there is no evidence that there is a difference  in
mean response due to irrigtion method.
        s.

Results  of the other ANOVA tests were that soil order was  not
signficant,  high  and low rainfall levels in the alfisols  was
not  significant,  but  thickness of the  unsaturated  zone  in
entisols  and ultisols was significant.  In addition, an  ANOVA
in  which ultisols and entisols were grouped and spodosols  and
alfisols  were grouped gave a significant F value.   Therefore,
from   these  analyses,  it  appears  that  the   statistically
different  mean  responses of the sixteen  scenarios  breakdown
into three groups;

    1)   entisols and ultisols with thin unsaturated zones
         (< 270 cm),

    2)   entisols and ultisols with thick unsaturated zones
         (> 270 cm), and

    3)   alfisols and spodosols.

Further  justification  for combining the results of  scenarios
is  given by the frequency plots of the annual pesticide  loads
from  each  scenario.  These are shown in Figures 4.13  through
4.16.   Figures 4.13 and 4.14 show the eight plots  for entisols
and  ultisols, thin and thick unsaturated zones, no irrigation
and  irrigation  practices  for each.  Notice that  the  curves
group   in both figures by depth of unsaturated zone.  The  thin
unsaturated  zones  fall virtually on top of one another.   The
thick   soils are also grouped but not to the extent of the thin
soils.
                              184

-------
DATA
           Irrigation Method

             N           O
          1.82E-2

          1.02E-4

          2.33E-2

          1.61E-3

          3.36E-3

          6.65E-3

          3.17E-4

          4.31E-4


ANOVA Table



Source

Irrigation method

Error

Total
                       4.24E-2

                       6.92E-4

                       7.59E-2

                       3.87E-3

                       3.56E-3

                       5.59E-3

                       4.45E-4

                       4.42E-4
Sums of
Squares
0.0004
0.006
0.0064
d.f .

1
14
15
Mean
Square
0.0004
0.0004

                                                          0.91
  o.io
              • 3'102
  Figure 4.12  One way analysis of variance of the effect of
               irrigation on method annual pesticide mass
               leached to the saturated zone  (kg/ha).
                             185

-------
  100-r
   90"
 U
 X

 »
 Q
 <
80"
   70' •
 S 60'
   40
 ui 30 • •
 O
 z

 I 20 +
 O
 UJ
 O
 oc
 01
 Q.
   10-
             30 FOOT CORE


       	 OVERHEAD IRRIGATON


       	NO IRRIGATION

             9 FOOT CORE


       	OVERHEAD IRRIGATION


       	 NO IRRIGATION
     H	1 I  II
                    1 I I I
                                                        1	1—M-
1.E -5      1.E -4       1.E -3       -I.E -2


                      PESTICIDE LOAD IN KG/HA
                                                  1.E -1
                                                          1..E -0
Figure  4.13   Frequency of annual quantity of pesticide
                leached to  the saturated zone from Entisol
                scenarios.
                                 186

-------
                                                    30 FOOT CORE
                                                     OVERHEAD IRRIGATON
                                              	NO IRRIGATION

                                                     6 FOOT CORE

                                                     OVERHEAD IRRIGATION

                                                     NO IRRIGATION
   1.E-5
1.E -4        1.E -3       1.E -2

          PESTICIDE LOAD IN KG/HA
1.E -1
1..E -0
Figure 4.14
 Frequency of  annual  quantity of  pesticide
 leached  to the saturated  zone  from Ultisol
 scenarios.
                               187

-------
  100
Q
111

§ 80
o
X
co
   80- •
O  .

s70

S  60
I-

2  so
   40-
tu 30 - •
O
z

x 20 +
O
   10-
o
cc
ui
Q.
   1.E-5
          -t—f-
               HIGH RAINFALL


              	  OVERHEAD IRRIGATION


              —  NO IRRIGATION

               LOW RAINFALL


              	  OVERHEAD IRRIGATION


              ....  NO IRRIGATION
                                -i—I I I I
              1.E -4       1-E -3       -I.E -2


                         PESTICIDE LOAD IN KG/HA
1.E -1
1.E -0
Figure 4.15   Frequency of annual quantity of  pesticide
                leached to  the saturated zone from the
                Spodosol scenarios.
                                 188

-------
                                          HIGH RAINFALL

                                     	 OVERHEAD IRRIGATION

                                     	NO IRRIGATION

                                          LOW RAINFALL

                                     	OVERHEAD IRRIGATION

                                     	 NO IRRIGATION
   1.E -5       1-E -4       1-E -3       1.E -2        1.E -1

                         PESTICIDE LOAD IN KG/HA
1.E -0
Figure 4.16  Frequency of annual quantity  of pesticide
               leached to  the saturated zone from the
               Alfisol scenarios.
                                189

-------
Figures  4.15  and 4.16 show the same frequency curves for  the
spodosols  and  alfisols.  In these plots, all the  curves  are
nearly  coincident.  These figures reveal that highest leaching
losses  were  less then 1 Kg/ha over the fourteen year  period.
The  frequency  of occurrence of any given level  of  pesticide
loading  to  the  saturated zone can also be  determined.   For
instance,   in   Figure   4.13  for  a   no-irrigation,   thick
unsaturated  zone  entisol scenario, there is  approximately  a
90%  chance  that the annual pesticide leach load in any  given
                             -4
year  will  exceed  1.0  x 10   Kg/ha.  There  is  only  a  10%
                                                             -2
chance,  however,  that  the annual load will exceed 1.0 x 10
Kg/ha.

By  condensing the information in the sixteen scenarios down to
three,  as mentioned earlier, better probability estimates  are
obtained.   Figures 4.17 through 4.19 show the three  resulting
frequency   distributions   for  the  thick  unsaturated   zone
ultisols  and  entisols,  thin unsaturated  zone  ultisols  and
entisols,  and  the alfisol and spodosol groupings.  Points  in
the  combined  data  sets were ordered and  plotting  positions
calculated using the Weibull distribution (Chow, 1964).

Of  course,  the loads shown in these figures are leached  only
from  the  treated band.  Much of the area in the  unit  citrus
block   is   untreated,   with  no  resulting   leached   load.
Therefore,  the  loads from Figures 4.17 through 4.19  must  be
reduced  by  the  ratio of treated to total area  in  the  unit
block.   This ratio is approximately 0.32 to 1 for ridge citrus
as  per  Figure 4.9, and 0.20 to 1 for double  bedded  flatwood
citrus  assuming a 3 m  (10 ft) band down the middle on a 7.6  m
by 4.6 m  (25 by 15 ft) tree spacing.

Table  4.5 summarizes the pesticide loads for the 90,  50 and 10
percentile   exceedance  probabilities  for  the  three   final

                             190

-------
                           RIDGE SOILS

                        THICK UNSATURATED ZONE
    1.E-5      1.E-4       1.E-3      1.E -2

                       PESTICIDE LOAD IN KG/HA
                                1.E-1
1.E 0
Figure  4.17
Frequency of annual  quantity of pesticide
leached  to the saturated zone from thick
Entisols and Ultisols.
                             191

-------
   0.00
     1.E-5
                          RIDGE SOILS

                       THIN UNSATURATED ZONE
1.E-4      1.E-3       1.E -2

         PESTICIDE LOAD IN KG/HA
1.E-1
1.E 0
Figure  4.18  Frequency of annual  quantity of pesticide
              leached  to the unsaturated zone from thin
              Entisols and Ultisols.
                              192

-------
                         FLATWOODS SOILS

                       SPODOSOLS AND ALFISOLS
     1.E-5       1.E-4      1.E-3       1.E -2

                         PESTICIDE LOAD IN KG/HA
1.E-1
1.E 0
Figure  4.19  Frequency of annual  quantity of pesticide
              leached to the saturated zone  from Spodosols
              and Alfisols.
                                193

-------
TABLE 4.5  SUMMARY OF PESTICIDE LOADINGS PER UNIT CITRUS BLOCK
           AREA FOR THE THREE FINAL UNSATURATED ZONE SCENARIOS
                                   Exceedance Probability
Scenario                       0.90         0.50         0.10
Alfisols and spodosols        1.8E-5       4.0E-4       2.2E-3


Thick ultisols and entisols   6.4E-6       4.2E-4       3.2E-3


Thin ultisols and entisols    1.9E-3       1.3E-2       9.6E-2
                              194

-------
unsaturated  zone scenarios.  These loads have been  multiplied
by  the  above  ratios  to yield loads per  unit  citrus  block
instead  of  unit treated area.  The table reveals that  lowest
loadings  at  all percentiles are associated with alfisols  and
spodosols.   For this scenario there is only a 10%  probability
that  the  pesticide load leached to ground water  will  exceed
0.002  Kg/ha.   Loads from thick entisols and  ultisols  exceed
those  for  the alfisols and spodosols slightly.   The  highest
loadings  emanate  from  thin  unsaturated  zone  entisols  and
ultisols.  There  is  a 10% chance that loads to  ground  water
under  these  soils  will exceed 0.1 Kg/ha.   Recall  that  the
thickness  of  the unsaturated zone in this scenario is 180  to
270  cm   (6 to 9 ft) and that the input load is 5.6 Kg/ha or  5
Ib/acre.   Increasing  or decreasing the load in any  of  these
scenarios  by  a  ratio  'x' would result  in  an  increase  or
decrease  in the leached load by the same ratio.  For instance,
if  the  application rate were doubled from 5.6 Kg/ha  to  11.2
Kg/ha,  the simulated load at the 10% exceedarice level in  thin
entisols  and  ultisols would also double, from 0.096 Kg/ha  to
0.19 Kg/ha.

Also  of  interest is the quantity of aldicarb and that of  its
two   toxic  metabolities  in  the  leached  load.   Table  4.6
summarizes  the average percentage of the load accounted for by
each  species.  Overall, very little aldicarb parent is leached
to  the   saturated  zone under any scenario.   Under  the  thin
unsaturated  zone  ultisols  and entisols, about  60%  aldicarb
sulfoxide  and  40% aldicarb sulfone makes up the leached  load
(a  1.5   to 1 ratio).  In the thick unsaturated  zone  entisols
and  ultisols,  the ratio is closer to 0.17 to 1, sulfoxide  to
sulfone.   The spodosols and alfisols, the ratio is roughly the
same, 0.19 to 1.

Obviously,   since   these  transformations   are   kinetically
controlled,   the   quantity   of  aldicarb   and   its   toxic

                             195

-------
TABLE 4.6  MEAN PERCENTAGE OF ALDICARB, ALDICARB SULFOXIDE AND
           ALDICARB SULFONE IN THE SIMULATED LEACHED
           PESTICIDE LOAD
Scenario
EN2.7
EN9.0
E02.7
E09.0
UNI. 8
UN9.0
U01.8
U09.0
SN1 . 2L
S01.2L
SN1 . 2H
S01.2H
AN1.2L
A01 . 2L
AN1.2H
A01 . 2H

Aldicarb
3
0
3
0
1
0
1
0
0
0
0
0
0
0
0
0
Percent Chemical
Sulfoxide
67
21
66
23
56
7
54
7
13
15
18
17
19
16
16
14

Sulfone
30
79
31
77
43
93
45
93
87
85
82
83
81
84
84
86
                              196

-------
metabolities  appearing  in the leachate is a function  of  the
residence  time  of  the chemical in the profile.   The  sooner
after  application  the pesticide is leached to  the  saturated
zone,  the more aldicarb and aldicarb sulfoxide will appear  in
the  leachate.   Figure 4.20 shows an example of the timing  of
the  mass  load  arrival at the saturated zone by  month.   The
quickest  breakthrough and highest percentages of sulfoxide are
associated   with   the   entisols  and  ultisols   with   thin
unsaturated  zones.   Over 90% of the total is leached  to  the
saturated  zone  within the first four months.  For  the  thick
ultisols  and  entisols, the breakthrough curve is  shifted  to
the  right,  however, the major portion of the  chemical  still
leaches  out  in  about  four months.  For  the  spodosols  and
alfisols,  the  shape of the breakthrough curve  is  different.
Chemical  leaches  from the soil move evenly in time for  these
soils.   Not  only does timing of the breakthrough  affect  the
quantities  of  aldicarb  and  its  daughter  products  in  the
leachate,  but  it  also has other important  implications  for
modeling  of  the saturated zone.  These are discussed  in  the
next section.
4.2  SATURATED ZONE MODELING
4.2.1  Choice of Model
The  Coupled  Fluid  Energy and Solute Transport   (CFEST)  Code
(Gupta  et  al., 1980) was selected to predict the movement  of
aldicarb  in the saturated zone.  CFEST is a multi-dimensional,
transient   or  steady  state,  saturated,  flow,  energy,  and
contaminant  transport  model  for predicting head  and  single
species  solute  concentration  in  a  confined  aquifer.   The
numerical   solution  scheme  used  is  the  standard  Galerkin
                              197

-------
Q
UJ

o

UJ


UJ
Q

5


tO
UJ
Q.

U.
O
UJ
U
oc
UJ
CL

UJ

P

_l

i

o
100-


 90-



 80-


 70-


 60-



 50-



 40-


 30-



 20-



 10-


 0
r
                            ULTISOLS, THIN

                            UNSATURATED ZONE


                            ULTISOLS , THICK

                            UNSATURATED ZONE



                           SPODOSOLS
                                  —(	1	1	r
        J   FMAMjjASOND

                        MONTHS
Figure 4.20   Cumulative mass  curve of pesticide  leaching
               within  the year  for three  representative
               scenarios.
                           198

-------
finite-element   method.   The  flow  system  may  be  complex,
multi-layered,   heterogeneous,  and  anisotropic,  with   time
varying  boundary conditions and time varying areal sources and
sinks.

Solute  transport  is calculated considering the  processes  of
dispersion,    diffusion,   and   convection.    Sorption   and
contaminant degradation can also be simulated by the model.

In  this  study, CFEST was used to model each aquifer  geometry
with  and without decay and with different pumping rates,  well
distances,  and well depths.  The model was run in two or three
dimensions,  depending on the aquifer geometry and well  depth.
The  concentration  of aldicarb in the drinking water well  was
predicted  by  CFEST using steady state ground-water flow  with
transient transport of aldicarb.
4.2.2  Model Configurations
Of  the  six aquifer systems delineated in Section 2.2.4,  four
were   simulated  with  variable  well  positions  arid   rates:
Floridan  worst  case, Floridan average cases, surficial  worst
cases,  and  two-aquifer  worst cases.  The  surficial  average
cases  and the two-aquifer average cases were not simulated due
to  preliminary hydraulic results indicating very slow  ground-
water  velocities  (0.002 m/day and 0.03 m/day in the  surfical
aquifer,  respectively).  At this rate it would take many years
for  water  to travel from the edge of the source area  to  the
drinking  water  well  91 m (300 ft) away.  Even with  a  decay
rate  two  or  three times slower than the rate  used  in  this
study,  no significant quantities of aldicarb would ever  reach
the  well.   Each aquifer system simulated with  the  saturated
model  was  run with two pumping rates, two distances from  the

                             199

-------
source  to  the  pumping well, and with and without  decay.   A
two-dimensional  X-Y or three-dimensional X-Y-Z simulation  was
used   for   each  flow  scenario.   The  two-dimensional   X-Y
simulation  was used to model the single aquifer configurations
when  the  drinking  water well fully penetrates  the  aquifer.
Because  the well itself integrates the aldicarb  concentration
over  the aquifer thickness, the two-dimensional X-Y simulation
which  integrates  over the vertical dimension is  appropriate.
The   three-   dimensional  X-Y-Z  simulation  is   used   when
simulating  the two-aquifer system and a partially  penetrating
well  in  an unconfined aquifer.  In all, a total of  44  model
simulations  were  performed.  The model input  parameters  for
all the cases are listed in Table 4.7.

A  plan  view  of the modeled region is shown in  Figure  4.21.
Although  the  region is 490 m  (1,600 ft) wide, only  half  the
width  was  simulated in the model since the other half is  the
mirror  image.   Figure 4.21 shows a distance of 91 m (300  ft)
between  the  source  and  pumping well.  Each  case  was  also
simulated  with  this distance being 300 m (1,000 ft).  Due  to
dimension  constraints  of  the CFEST code, it was  not  always
possible  to represent the field  (left to right in Figure 4.21)
as  being  825  m (2,700 ft) wide.  Widths of 550 m and  275  m
were  also  simulated  in some of the cases as shown  in  Table
4.7.

These  different boundary conditions were used in the  scenario
simulations  as  shown  on the view of the model  area  (Figure
4.21).   The boundaries parallel to the direction of flow  were
no-flow  boundaries.  The boundary up-gradient of the well  was
a  constant-head  or a specified  flux boundary.  The  scenarios
which  simulated  partially  penetrating shallow wells  in  the
Floridan  Aquifer used constant-head up-gradient boundary.  All
the  other  scenarios  used a specified flux boundary  for  the
up-gradient  boundary  that maintained the  regional  gradient.

                             200

-------
    TABLE 4.7  MODEL INPUT  PARAMETERS  FOR THE SIX AQUIFER TYPES SIMULATED












NJ
O
H1







Aquifer Type
Floridan Worst Case
(Shallow Well)
Floridan Worst Case
(Deep Well)
Floridan Average Case
(Shallow Well)
Floridan Average Case
(Deep Well)

Surftclal Worst Case
(Shallow Well)
Two-Aquifer Case
1) Surflclal
2) Confining Layer
3) Floridan

Field
Width
(ft)

2.700

2,700

1,800

1,800


900

900
-


Aquifer
Thickness
(ft)

600

350

600

350


35

40
10
400
Well
Screen
Interval
(ft)

10-40

0-350

10-40

0-350


0-35

.
-
50-400

Hydraulic
Conductivity
(ft/day)

5.000

5,000

1.000

1,000


100

40
0.025
1,770


Porosity
(X)

20

20

30

30


15

20
5
20


Gradient
(ft/ml)

3

3

2

2


4

25
40
10


Recharge
(in. /year)

25

25

13

13


30

20
-



Model
Dimension

3D

20

3D

2D


2D

30
-

Model
Grid
Spacing
(ft)

150

150

40

40


15

30
-

Mode
Tiim
Scei
(day

10

10

30

30


30

30
-

(1) Vertical Gradient

1 ft  =  0.3048 m
1 in  =  2.54 cm

-------
                                        NO FLOW BOUNDARY
to
o
                                         2700

                                        NO FLOW BOUNDARY


                                               Dimensions In Ft
           Figure 4.21  Plan view of a hypothetical citrus grove  and well configuration.


                         1 ft =  0.3048 m

-------
The  boundary down-gradient of the well were all  constant-head
boundaries.

                                                     -5    ,  j
The  annual loading rate used in the model was 1x10   gm/cnr.
This  loading rate was converted to an input concentration  for
each  aquifer  case based on the average annual recharge.   The
input  concentrations  at the source area (field)  ranged  from
399 to 920 ug/L (see Table 4.8).

All   cases  were  simulated  with  and  without  decay.    The
half-life  used  in the model was 29.7 days.  Nine  meters  and
1.8   meters   were  used  for  longitudinal   and   transverse
dispersion,   respectively.   Typically one-tenth  the  critical
distance  of  travel  is used for longitudinal  dispersion  and
transverse  dispersion  is taken to be one fifth the  value  of
longitudinal  dispersion.  The critical distance from the  edge
of  the field to the well is 91 m (300 ft), making longitudinal
dispersion  9 m and the transverse 1.8m.  In general the model
is  not  very  sensitive to dispersion, so when  the  well  was
simulated  at  300 m (1000 ft) from the source area,  the  same
dispersion values were used.
4.2.3  Input from Unsaturated Model
The  output  from the unsaturated zone model PRZM is the  daily
water  flux  (cm) and daily pesticide loadings  (gm/cm  ).   This
one-dimensional,  daily  pesticide output must  be  transformed
into  meaningful input data for a single chemical component  in
a  two or three-dimensional ground-water model with a  10 to  30
day  timestep.  There are three aspects to interfacing  the  two
models:
                              203

-------
    1)   determining the magnitude of the pesticide load to
         the saturated zone;

    2)   determining the temporal distribution of this load;
         and

    3)   determining the spacial distribution of the load.

The   magnitude  and  timing  of  the  pesticide  load  to  the
saturated  zone  would typically be prescribed directly by  the
output  from PRZM.  Unfortunately, due to the time  constraints
of  the  study,  the  two  modeling  efforts  had  to  be  done
simultaneously.    A  standard  unit  pesticide  load  of  1  x
10~5  gm/cm2 (1  Kg/ha)  was  used  for  all  the   scenarios.
Preliminary  results  from  the  unsaturated  zone  simulations
showed  that this load was within the range of expected values.
Since  the output of CFEST is linear with respect to input load
(see  Section  4.2.6.2), the actual well  water  concentrations
for  a  given  combination of saturated  and  unsaturated  zone
scenarios  can be determined by a multiplicaton factor based on
the  actual output pesticide load from PRZM and the unit  load.
Therefore,  the  output of CFEST is expressed as  a  "relative"
concentration,  that  is,  the concentration simulated  at  the
well  divided by the initial concentration calculated from  the
unit pesticide load and the recharge rate shown in Table 4.8.

With  pesticide applications occurring every year, there was  a
possibility  of  aldicarb  accummulating in the  soil  and  the
loads  to the saturated zone increasing.  This possibility  was
tested  by  analyzing the total annual leached load  from  each
year  simulated.  If the loads measured in each successive year
showed  an  increasing  trend, this  would  indicate  pesticide
accummulation.   The  results  of analyzing  the  annual  loads
showed  that  there  is  no upward trend of  loads  during  the
simulation period.

                              204

-------
TABLE 4.8  INITIAL INPUT CONCENTRATIONS USED IN CFEST FOR EACH
           AQUIFER SYSTEM SIMULATED (in ppb)
           Aquifer        Recharge       Concentration
           System	(in/yr)	(ppb)	
       Floridan, worst       25               479

       Floridan, average     13               920

       Surficial, worst      30               399

       2-Aquifer worst       20               599


(1 inch = 2.54 cm)
                             205

-------
The  second  aspect in linking the unsaturated zone  output  to
the  saturated  zone  modeling  effort is  determining  how  to
distribute  the pesticide loads in time.  Because of the manner
in  which  the timestep is handeled in CFEST,  daily  pesticide
loads  had to be aggregated to monthly loads.  Average  monthly
loads  were  calculated  to help determine the best  method  to
distribute  the  annual  unit load.  These monthly  loads  were
summed  to determine the number of consecutive months  required
for  over 90% of the pesticide to be leached from the saturated
zone.   Table 4.9  shows the results of this analysis for  each
scenario.

For  the  ultisols and entisols, over 90% of the  pesticide  is
leached  out  within  four and one-half months.   In  February,
pesticide  would  only  be leached out the second half  of  the
month  after the pesticide application in mid-February.  In the
thick  entisols and ultisols 5 to 6 months was usually required
to   leach  out  over  90%  of  the  pesticide.   It  is   also
interesting  to  note  that the primary  leaching  months  over
which  leaching  occur start later in the year, reflecting  the
additional  time  required for leaching through a soil four  to
five times as thick.

The  alfisols in areas with high rainfall leach close to 90% of
the  pesticide  within four months with irrigation.   In  areas
with  lower rainfall, irrigation enhances the leaching so  that
most  of  the pesticide is removed within four or five  months.
Without  irrigation,  seven  months is required to  remove  the
same amount.

The  spodosols lengthen the time required to account for 90% of
the  leachate  to  between  five and  eight  months.   The  low
permeability   layer  in  the  soil  slows  down  and  prolongs
leaching.   As  with the alfisols in areas with high  rainfall,

                              206

-------
    TABLE 4.9  TIME REQUIRED  TO LEACH 90% OF THE PESTICIDE FROM THE UNSATURATED ZONE
                                          Consecutive
to
o




Soil Type
ULTISOLS



ENTISOLS



SPODOSOLS



ALFISOLS






Irrigation
Method
6 FT None
Overhead
30 FT None
Overhead
9 FT None
Overhead
30 FT None
Overhead
LOW RAIN None
Overhead
HIGH RAIN None
Overhead
LOW RAIN None
Overhead
HIGH RAIN None
Overhead
Months
Necessary
to Leach
90% of
Pesticide
5
4
4
5
5
4
6
6
8
7
6
6
7
5
5
4


Months
of
Leaching
Feb-June
March-June
June-Sept
May-Sept
Feb-June
Feb-May
March-Aug
March-Aug
March-Oct
March-Sept
April-Sept
March-Aug
March-Sept
April-Aug
April-Sept
April-Aug



Percent
Leached
95
93
94
90
99
97
92
94
96
93
93
93
95
90
95
90


Percent
Leached in
4 Months
86
93
94
86
88
97
78
58
67
58
78
81
78
84
89
90

-------
the  pesticide travels faster requiring six months versus seven
to eight months in areas with low rainfall.

The  more  time the leaching of the annual load  requires,  the
smaller  any load to the saturated zone will be.  On the  other
hand,  if  the  entire annual load was leached in  one  or  two
months,  the  concentrations entering the saturated zone  would
be much higher.

The  average of the number of months necessary to leach 90%  of
the  pfe-sticide is just over five.  A four month period was used
in  the  simulations as this decision had to be made  a  priori
based  on limited simulation results.  A four month period  for
loading  will  result in slightly higher concentrations at  the
well  for  scenarios  in which the load is distributed  over  a
longer  period of time.  The last column of Table 4.9 shows the
percentage of the pesticide that is leached in four months.

The  spatial  distribution  of  the pesticide load  is  a  more
complex  problem, which effects both the peak concentration  at
the  well  and  the time necessary for the peak  to  reach  the
well.   For some scenarios it was discovered that the width  of
the  source  area had an effect on  well-water  concentrations.
Since  actual  field  sizes in Florida vary  tremendously,  the
following  analysis  was done in an attempt to understand  this
relationship    so   model   results   could   be   interpreted
accordingly.
4.2.4  Source Area and Well Concentration
A  plug  flow analysis  reveals  the basic   relationship  between
the  field  width and the  concentration at the well.   The  key
variable  is the residence time under  the  field which  combines

                              208

-------
the  ground-water velocity associated with a given scenario and
the  width of the field.  As the residence time under the field
increases,  the  water  may be subjected to  several  pesticide
loading  events.  In cases with no decay or low decay, this can
result  in higher concentrations at the well.  Figure 4.22 is a
schematic  graph of what the concentration of the water at  the
well  would  be with and without decay based on  the  residence
time   under  the  source  area.   At  the  beginning  of   the
simulation  the  concentration  at the well will be  zero.   It
will  remain zero for the length of time it takes the water  to
travel  from the leading edge of the field to the well.   After
that,  the  concentration in the case of plug flow is  directly
related  to  the  residence time under the  source  area.   The
concentration  increases as more and more pesticide is  leached
to  the  ground  water.   At the end of four  months,  no  more
pesticide  is  leached  from  the unsaturated  zone  until  the
following  year.  This means that even if the residence time is
over  four  months,  the  water only receives  four  months  of
pesticide  loading  in  one year.  One way  to  visualize  this
process  is to imagine a parcel of water moving under the field
(Figure  4.23).   While it is under the field and  leaching  is
occurring,  the concentration in the parcel of water  continues
to  rise  at  a steady rate.  If the water is under  the  field
when  leaching  is not occurring or when it is  traveling  from
the  field edge to the well, the concentration remains constant
(assuming zero decay).

When  there  is  no  decay the water reaches  a  local  maximum
concentration  at  a  residence time of four  months  and  then
plateaus  for  the eight months of no pesticide loading.   When
leaching  begins  again  the following year  the  concentration
begins  to  rise in the water below the field.  For longer  and
longer  residence times the concentration will continue to rise
following  this  general pattern.  In the case with  decay  the
pesticide  reaches  a  maximum in four months, then  begins  to

                              209

-------
g
H
CC
J-
UJ
o
o
o
                                 PLUG FLOW WITH NO DECAY
                                   PLUG FLOW WITH DECAY
                                      AND CARRY OVER
    —4,
PLUG FLOW WITH DECAY
 AND NO CARRY OVER
     TRAVEL
     TIME
    TO WELL
  4 MONTHS     8      1 YEAR

          RESIDENCE TIME UNDER FIELD
                                                           2 YEARS
Figure  4.22
Schematic  graph  of pesticide  concentration
through  time based on residence time under
source area.
                                210

-------
                                                   r	C
   DRINKING
    WATER
\   WELL

-1_
                'PARCEL* OF WATER
^^:^y^y^^^'::-:^^'i^f^.
Figure 4.23  Schematic  diagram of pesticide accumulation
              process.
                           211

-------
decay  during  the  eight  months of  no  loading.   The  cycle
repeats itself annually.

For  a given residence time there are two general patterns:  1)
when  the  residence  time is less than one year,  and  2)  the
other  is  when  the residence time is greater than  one  year.
Figure  4.24  shows what the concentration pattern  would  look
like  if  the  residence time of ground water under  the  field
were  eight months.  The concentration would rise to a  maximum
during  the four initial months of leaching, maintain that peak
for  four months and then decrease for four months.  The  cycle
then  repeats  itself  year after year.   For  residence  times
greater  than one year the concentration level will follow  the
same  pattern  shown  in  Figure  4.22,  leveling  off  at  the
concentration corresponding to the appropriate residence time.

The  decay curve has basically the same shape regardless of the
residence  time.   It is conceivable that if all the  pesticide
does  not  decay  in one year, the  annual  peak  concentration
could  go  up over the years as shown in the curve with  carry-
over in Figure 4.22.

The  very  real process of dispersion requires modification  of
the  plug flow theory.  Dispersion causes the spreading of  the
contaminant.   Figure  4.25 shows how dispersion  would  change
the  concentration  in  the well predicted by plug flow  in  an
example where the residence time was greater than three years.

A  series  of sensitivity runs were made with CFEST for two  of
the  aquifer  cases to help understand this  relationship.   In
the  first  set  of sensitivity runs the  Floridan  worst  case
deep  well,  1,000 gpm pumping) was run both with  and  without
decay  where  the field width ranged from 46 m (150 ft)  up  to
823  m  (2,700 ft)  (the maximum field width for this case).   A
second  set  of  sensitivity runs was made  for  the  surficial

                             212

-------
jjji
z
o
Ul
O
z
o
o
   —1r
      0
  TRAVEL

  TIME

 TO WELL
   PLUG FLOW. NO DECAY. RESIDENCE TIME = 8 MONTHS
                2YR           3YR



TIME (AFTER FIRST PESTICIDE ARRIVAL)
4YR
Figure  4.24  Concentration profile  for eight  month

              residence  time based on plug  flow.
                           213

-------
 Z
 UJ
 o
 z
 o
 o
                         PLUG FLOW-
                                     WITH DISPERSION
  TRAVEL
  TIME

  TO WELL
TIME (YEARS)
Figure  4.25  Effect of  dispersion on concentration in the

              well over  time.
                             214

-------
worst  case (shallow well, 200 gpm pumping rate) both with  and
without  decay where the field width ranged from 46 m (150  ft)
up  to 275 m (900 ft) (the maximum field width for this  case).
In  all  scenarios the pumping well was 91 m (300 ft) from  the
edge  of the field.  Each sensitivity simulation ran for  three
years  with  identical  model parameters except for  the  field
width.

The  results of the sensitivity analysis for the worst case  of
the  Floridan  Aquifer  are  shown in  Figures  4.26  to  4.28.
Figure  4.26  shows the concentration at the well  versus  time
for  the first year of simulation with no decay.  This  clearly
demonstrates  how  the concentration at the well  increases  as
the  field  width increases.  Figure 4.27 is the same  scenario
with  decay.   There the concentration also increases with  the
field  width  to  a critical width of 366 m  (1200  ft)  beyond
which  the  concentrations essentially do not  change.   Figure
4.28  summarizes  these trends for all the scenarios  with  and
without decay.

The  results  for  the  surficial worst cases,  both  with  and
without  decay,  are shown in Figures 4.29 and 4.30.   For  the
case  without decay (Figure 4.29), a peak concentration at  the
well  was  never  reached in the three  year  simulation.   The
relationship  between  the  maximum concentration  after  three
years  and field width is such that the concentration gradually
increases  as  field  width  increases.  The  case  with  decay
(Figure  4.30)  does reach a peak concentration at the well  on
an  annual cycle.  This peak does not change as the field width
changes.

In   all  these  simulations,  the  fields  modeled  all   have
residence  times much greater than four months.  As observed in
the  Floridan worst case sensitivity runs, for residence  times
greater  than  a critical time of about three months, the  peak

                             215

-------
                   ••\\.     \     \
                                                  A 150 ft


                                                  X 300 ft
           • £50 «_


           D 600 ft


           B 1200_ft


           * 1800 ft


           X 2700 ft
              4 MONTHS      8 MONTHS

                    TIME, MONTHS
12 MONTHS
Figure 4.26   Concentration histories  at variable field
              widths for the Floridan  worst case scenario
              with no decay.
                           216

-------
     1 -I
o
LU
O
z
o
o
                         1200 & 1800
                 4 MONTHS      8 MONTHS


                       TIME, MONTHS
12 MONTHS
           A 150 ft



           X 300 ft
           D 60_0_ft



           H 1200ft



           4 1800ft



           X 2700 ft
Figure  4.27  Concentration histories at  variable field widths

              for  the Floridan worst case scenario with decay.
                              217

-------
to
M
00
               D)
                  4-
                  3H
CE
               LU  2-
               O
               O
               CJ
               UJ
               Q.
                  1-
                                                  Legend
                                                  A pjormxn MB
                          500    1000   1500   2000   2500
                              FIELD  WIDTH,   Ft
                                             3000
           Figure 4.28  Relationship of  field width to peak concentration at the
                        well for the Floridan worst case.

-------
   30-
 09

 i
 P
 tt 20-
 ui
 o
 o
 u
   10-
      0    1 YEAR       2 YEARS       3 YEARS

                     TIME, YEARS
  LEGEND
A150 ft


X 300 ft
                                                    D 600 tt


                                                    B 900 ft
Figure 4.29  Concentration  histories at variable field widths
              for the surficial worst case scenario without
              decay.
                             219

-------
    1.5-
O


oc
 UJ
 u

 O
 O
                  150ft

                  300ft

                  600ft

                  900 ft
                4 MONTHS       8 MONTHS


                     TIME, MONTHS
                             12 MONTHS
Figure  4.30
Concentration histories  at variable
field widths for the surficial worst
case scenario with decay.
                      220

-------
concentrations are the same.

The  two cases run in this sensitivity analysis range over  the
spectrum  from the fastest travel time (Floridan worst case) to
the slowest  travel  time (surficial worst case) of  the  cases
simulated.   The relationship between peak concentration at the
well  and  field  width for the Floridan  average  cases  would
probably fall between the two cases studied.

Based  on  these  results it appears that, for all  cases  with
decay,  the  peak concentration at the well is  independent  of
field  width  except  for small widths (i.e.,  residence  times
less  than four months).  For the cases simulated without decay
the  maximum  concentration is dependant on the residence  time
of  the water under the field.  The maximum concentration  will
be  higher  for  wider fields.  The exact maximum  depends  not
only  on  the field width but the ground-water velocity  (i.e.,
travel time to the well) and dispersion.

It  should  be noted here that the regularity of  the  cyclical
and  "rising and plateauing" well-water concentration  patterns
shown  in the next section is caused by the loading assumptions
made  for  interfacing  the two models and  the  assumption  of
steady  flow.   While in nature, these patterns will be  highly
irregular  and noisy, the same general trends discussed in  the
previous section should emerge.
4.2.5  Results of Aldicarb Simulations
A  summary  of  the  results for  the  four  simulated  aquifer
geometries  are  shown  in Tables 4.10 to 4.13.   These  tables
show  the relative peak concentrations, the time of arrival  of
the  peak  along with the other parameters that  describe  each

                             221

-------
specific  scenario.   In  addition to these  tables,  for  each
scenario,  a  graph  is  presented showing  the  trend  of  the
relative  concentration  at  the  well during  the  three  year
simulation  period (Figures 4.32 to 4.39).  Four scenarios  are
shown  on each graph for ease of comparison.  For a given  well
distance  (i.e., 91 m versus 300 m) and well depth  (shallow  or
deep),  concentration histories are shown for the high and  low
pumping rates, with and without decay.

The  results  for the worst case for the Floridan  Aquifer  are
shown  in Figures 4.31 and 4.32 and Table 4.10.  The worst case
for   the   Floridan  Aquifer  has  the  highest   ground-water
velocities  (4.3  m/day  or 14.2 ft/day based on  the  regional
gradient).   With  such high velocities, the ground  water  has
only  a  six-month residence time, the shortest residence  time
of  all  the  scenarios.  This short residence time  result  in
cyclical  concentration  histories  for the  scenarios  without
decay  as well as the scenarios with decay, as shown in Figures
4.31  and 4.32.  Recall that with residence times less than one
year,  the  concentration  histories is cyclical  in  scenarios
with  no  decay  (see  Figure  4.24).   Considering  the  short
residence  time  the  relative concentrations at the  well  are
high  in comparison to most other scenarios.  This is also  due
to  the  high velocities.  There is little time for  dispersion
to  spread out the contaminant over the distance from the grove
(source  area)  to  the  drinking, water  well.   Based  on  the
well-induced  velocities,  the  ground water takes  from  15-20
days  to  travel to the well 91 m from the source  and  62.5-69
days to travel to the 300 m well.

The  average  cases  for the Floridan Aquifer  (Table  4.11  and
Figure  4.33-4.36)  fall  into two  catagories,  the  scenarios
without  decay and scenarios with decay.  In the scenarios with
decay,  the concentration at the well cycles annually, as shown
most  clearly on Figures 4.34 and 4.36.  The annual peak occurs

                             222

-------
TABLE 4.10  SUMMARY OF RESULTS FOR FLORIDAN WORST CASE
            SIMULATIONS
Well
Depth
(ft)
10-40
10-40
10-40
10-40
10-40
10-40
10-40
10-40
0-350
0-350
0-350
0-350
0-350
0-350
0-350
0-350
Well
Distance
(ft)
300
300
300
300
1,000
1,000
1,000
1,000
300
300
300
300
1,000
1,000
1,000
1,000


Decay
No
No
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
Pumping
Rate
(gpm)
200
40
200
40
200
40
200
40
1,000
500
1,000
500
1,000
500
1,000
500
Peak
Relative
Concentration
.1.5 x 10"2
1.6 x 10~2
3.1 x 10"3
3.0 x 10"3
1.3 x 10~2
1.3 x 10r2
7.3 x 10"4
7.1 x 10"4
9.0 x 10"3
9.0 x 10"3
1.9 x 10"3
1.9 x 10"3
8.7 x 10"3
8.7 x 10"3
7.0 x 10"4
6.8 x 10"4
Time
to Peak
(days)
150
160
120
120
230
230
150
150
150
150
120
120
210
210
140
140
                           223

-------
to
N)
       R
       8
  w
2S-
2d'
1-1
§§
53
O O

"ij
>• °.-1
•-> a
£
FVD(0)W(300.10-40.40)
   FWD(0)W(300,10-40.200)
                                A
                                              .        .
                                      w
                                      \
               FWO(29.7)W(300.10-40.200)
               FWD(29.7)U(300,10-40.40)
                  I         V
                             v    /     V
        0.0    180.0     360.0     540.0    730.0
                            TIME IN DfflfS

                    a.    91  m  to  well
                                  900.0
                                          1080.0
                                                                      °
                                                    Is
                                                                      i.
                                                                      a

                                                                      §
                                                                                 FUD(0)U(1000.10-40.40)
                                                                                 I    nro(0)H(1000.10-40.200)
                                                                    FVD(29.7)K(1000,10-40.40)
                                                                    FUD(29.7)W(1000.10-40.200)

                                                                  ^J*         .^•••'••x.
                                                         0.0     180.0     360.0     540.0     730.0
                                                                            TIME IN DflYS

                                                                   b.    300  ra  to  well
                                                                                                           900.0    1080.0
       Figure 4.31    Concentration versus  time for  Floridan worst  case  with  a  shallow
                          well.

-------
to
                     FWD(0)W(300.50-350.1000)
                     FUD(0)U(300.50-350.
              . FUD(29.7)W(300.50-350.1000)
             r FWD(29.7)W(300.50-350.500)
               180.0     360.0     540.0    730.0
                            TIME IN DRYS

                   a.    91  m  to  well
                                                 800.0     1080.0
                                                                    cu V'
                                                                    »•—t

                                                                      °
                FUD(0)W(1000.50-350.1000)
                FWD(0)W(1000.50-350.500)
                                                                                  FWD(29.7)U(1000.50-350.1000)
                                                                                  FMD(29.7)WMOOO. 50-350.500)
0.0     180.0    360.0     540.0     730.0
                    TIME IN DfttS

           b.    300  m to  well
800.0    1080.0
     Figure  4.32   Concentration versus  time  for  the  Floridan worst  case  with  a  deep  well

-------
  TABLE 4.11  SUMMARY OF RESULTS FOR FLORIDAN AVERAGE CASE
Well
Depth
JftL
10-40
10-40
10-40
10-40
10-40
10-40
10-40
10-40
0-350
0-350
0-350
0-350
0-350
0-350
0-350
0-350
Well
Distance
(ft)
300
300
300
300
1,000
1,000
1,000
1,000
300
300
300
300
1,000
1,000
1,000
1,000
Decay
No
No
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
Pumping
Rate
(gpm)
200
40
200
40
200
40
200
40
1,000
500
1,000
500
1,000
500
1,000
500
Peak
Relative
Concentration
8.3 x 10"3
8.8 x 10"3
5.1 x 10"5
4.6 x 10"5
2.5 x 10"3
2.4 x 10"3
1.6 x 10'8
1.2 x 10"8
7.4 x 10"3
8.8 x 10"3
5.1 x 10"5
4.5 x 10"5
3.6 x 10"3
4.3 x 10"3
5.5 x 10'8
4.2 x 10"8
Time
to Peak
(days)
A
A
150
150
A
A
300
300
A
A
120
150
A
A
270
300
Aj  Peak concentration was not reached  in  3-year  simulation.
                            226

-------
     00
NJ
     F-i 10
     
-------
       O
       a-
       o
       •ji-
       o
       *
NJ
M
00
    So.
    Of o
                 FAD(29.7)W(300.10-40.200)
   n
in
 •
K-
FAD('29.7JW(300.10-40,40)
Vff   i
                       g2
                                                                  to
        0.0     180.C     360.C    &U.O    723.0
                           TIME IN DRYS

                    a.   91 m to well
                                               9CC.O
                                                       108C.O
                                     FAO(29.7)H(1000.10-40.200)

                                     V
                                                                                               FAD(29.7)U(1000.10-40.40)
                           0.0     180.0     360.0     StO.O     720.0
                                              TIME IN DflYS

                                      b.    300  m  to  well
                                         900.0
                                                                          1C80.0
      Figure  4.34   Detail of Floridan  average  case with a shallow well  and decay.

-------
     2 2
     X °»
Ni
to
VO
              FAD(0)W(300.50-350.500)_ /'
                                        .50-350.1000)
                               ^FAD(Z9.7)W( 300.50-350.1000)
                               XFAD(29.7)W( 300.50-350.500)
        0.0    180.0     360.0     Stt.O    7X.O    900.0    1080.0
                            TIME IN DflYS
                   a.   91 m  to  well
                                                                    oq
                                                                   E*
                                                                                           FAD(0)H(1000.50-350,SCO)^ /
                                                                                       FAD(0)W(1000,50-350.1000
                            FAD(29.7)W(1000.50-350.1000)
                            FAD(29.7)W(1000.50-350.500)
0.0     180.0     360.0    640.0     730.0
                   TIME IN DflYS
          b.    300  m to well
                                                                                                               900.0    1080.0
        Figure 4.35    Concentration versus  time  for  the Floridan  average  case  with  a  deep
                          well.

-------
to
u>
o
        a-

        a

        «
aiS-l
t->


|3H

go

uKJ-
>
2-


o
                   FAD(29.7)U(300,50-350.1000)
                                                                   T

                                                                    2 2
                                                                    X «°
                                                                   a:
                                                                   a:

                                                           to


                                                           *—f



                                                             0
                                                                                FAD(29.7)W(1000.50-350.1000)
                                                                                    FAD(29.7)W{1000.50-350.500)
          0.0     180.0     360.0     5-10.0     720.0

                             TIME IN DflYS



                     a.    91 m to well
                                                 9CO.O
                                                         1C80.0
                                                               0.0     180.0     360.0     S-UJ.O     730.0

                                                                                  TIME IN DflYS



                                                                          b.    300  m  to  well
                                                                                                              900.0
                                                                                                                      iceo.o
             Figure 4.36    Detail  of  Floridan average case  with  a  deep well  and decay.

-------
anywhere  from  120  days to 300 days after  the  beginning  of
leaching,  depending  on the aquifer scenario and  distance  to
the  well.   In the cases without decay, the concentrations  at
the  well increase throughout the three-year simulation  period
with  minor  fluctuations (Figures 4.33 and 4.35).  The  ground
water  in  these average cases for the Floridan Aquifer  has  a
residence  time  of  almost four years.  With  such  a  lengthy
residence  time,  the  concentrations would  continue  to  rise
according  to the theory developed earlier on residence  times.
The  fluctuations  visible in the cases with the well at  91  m
are  a dispersed form of the plateaus predicted by a plug  flow
analysis.   At 300 m, these fluctuations have been smoothed out
by  dispersion.  The initial rise is also delayed, as would  be
expected when the well is further from the source area.

The  worst case for the surficial aquifer geometry shows trends
very  similar to the Floridan average cases because of its long
residence  time  of five years.  Table 4.12 shows the  relative
peak   concentrations  and  Figures  4.37  and  4.38  show  the
concentration   histories.   The  scenarios  with  decay  cycle
annually.   The relative concentrations are the highest of  all
of  the  scenarios with decay.  This is because the aquifer  is
very  shallow so the dilution of the contaminant can not be  as
great  as the systems with deep aquifers.  In the cases without
decay,  the  concentrations  continuously rise  throughout  the
three-year  simulation.   In these scenarios there is  a  large
difference  between  the high pumping rate and the low  pumping
rate.    In  the  surfical  aquifer  the  well  influences  the
regional  hydraulic  gradient  more than in any  of  the  other
aquifer  configurations.   The well with the high pumping  rate
is  able  to pump a higher proportion of clean water  than  the
well  with the lower rate.  The plateaus are much more defined,
to  the point of decreasing before the next increase.  This  is
due  to  the slow velocities and dispersion of the  contaminant
at the well.

                             231

-------
TABLE 4.12  SUMMARY OF RESULTS FOR THE SURFICIAL WORST CASE
Well
Depth
JftI
0-35
0-35
0-35
0-35
0-35
0-35
0-35
0-35
Well
Distance
(ft)
300
300
300
300
1,000
1,000
1,000
1,000
Decay
No
No
Yes
Yes
No
No
Yes
Yes
Pumping
Rate
(gpm)
200
40
200
40
200
40
200
40
Peak
Relative
Concentration
8.5 x 10"2
2.4 x 10"1
3.8 x 10"3
3.4 x 10"3
7.1 x 10~2
1.2 x Kf1
3.8 x 10"5
3.3 x 10"5
Time
to Peak
(days)
A
A
120
120
A
A
210
240
AJ  Peak concentration was not reached in 3-year  simulation.
                             232

-------
       R
       R
       8
NJ
u>
u>
00

u
     Ul
     a:.
                                     SWD{0)H(300.10-35.40)
                                          SW)(0)W(300.10-35.200)
                                 SKD(29.7)H(300.10-35.200)
                                'SVD(29.7)H(300.10-35.40)
         0.0     190.0     360.0     Stf.O     730.0

                             TIME IN DflYS
800.0
                                                          1080.0
                    28
                    So

                    SJ
                    *—I

                    Si
                    UO



                      8

                                                                                                            //ksHD(0)U( 1000.10-35.40)
                                                                                                       •SUD(0)U(1000.10-35.200)
                                                SWD(29.7)W(1000.10-35.200)
                                                . SUD(29.7)W(1000.10-35.40)
                                                                     0.0     180.0     360.0     540.0     730.0

                                                                                         TIME IN DfttS
                                                                                                                  900.0     1080.0
                     a.    91  m  to  well
                                    b.    300  m  to well
           Figure  4.37   Concentration versus  time  for  the surficial worst  case.

-------
* °~\ SKD(29.7)W(1000.10-35.200)
A*
B- f\
,4
• ' \
is~ // \
•— .• i i
t-« o s ' !|
(E • i ' <>
?? B- I ' t>
i / / \\



O 9- / i Vi
C3 " ! > li
: i : i
S«8 // \\

1 2" / 1 \\
~5 ! i : i
=-i » i •. i
ty . it : >
of : i •. i
O j ' i 't
2- // yv
°. //
u> :»
° /''
ra >
A A
// \ // \
// \ // \

/ / '\ / / M
// n /; V.
; ' ••• : t :>
/ ' Sl ; I '.'
/ \\ // \\
// \\ // \\
/ / \\ l> \\
// \\ // \\
// \\ // \
II u // i
SUD(29.7)W(1000.10-35.40)

               0.0   180.0    3EO.O    540.0   720.0
                               TIME IN DRYS
                                               9CO.O
                                       1CSO.O
Figure 4.38
Detail of the  surficial worst  case with the
well  1000 ft  from the  source with decay.
                              234

-------
TABLE 4.13  SUMMARY OF RESULTS FOR TWO-AQUIFER SYSTEM
            (WORST CASE)
Well
Depth
(ft)
50-400
50-400
50-400
50-400
Well
Distance
(ft)
300
300
300
300
Decay
No
No
Yes
Yes
Pumping
Rate
(qpm)
1,500
700
1,500
700
Peak
Relative
Concentration
1.8 x 10"3
2.2 x 10"3
3.2 x 10"4
3.0 x 10"4
Time
to Peak
(days)
150
150
120
120
                         235

-------
               Oo


               S ">'
               5-5
               LJ „•
               0(3
               U O'
               >• •*
               f^




               i5'
                                      r\
                         \*2W>{0)K(300.50-400.700)
                  0.0    180.0    360.0    540.0   720.0

                                  TIME IN DflYS
800.0
                                                          1080.0
Figure 4.39  Concentration versus time  for  the two-aquifer

               worst  case with the  well  300 ft from the  source,
                                 236

-------
Some  of the scenarios for the two-aquifer, worst case geometry
were  modeled.   In  these  cases the well  was  in  the  lower
aquifer.   Consequently the contaminant must travel through the
upper  aquifer  and  confining layer to reach  the  well.   The
boundary  conditions  for the two-aquifer cases were such  that
unusually  large concentrations of the contaminant accummulated
in  center  portion of the upper aquifer.  This led  to  higher
than  normal transport of the contaminant to the lower  aquifer
and  drinking water well.  The results are shown, nevertheless,
in  Figure 4.39 and Table 4.13.  Even with the hydraulics  that
enhance transport the relative concentrations are small.

Plug Flow Modeling Results—

A  simple  plug  flow model of the single aquifer  systems  was
also  used  to  aid in the understanding and  extrapolation  of
results  of the CFEST simulations.  Maximum concentrations  for
several  scenarios  were calculated using the plug  flow  model
developed  earlier in this section.  For these calculations the
contaminant  load was assumed to disperse evenly in a column of
water   (1  cm  x 1 cm x aquifer depth x porosity).   The  water
                                        -5      2
receives  the  annual  unit load (1 x 10   gr/cm  )  over  four
months.   In the cases with no decay this maximum concentration
does  not change while traveling to the well because of the  no
dispersion  assumption  in plug flow.  In the cases with  decay
the   concentration  was  first  calculated  under  the   field
assuming four months of loading:
             C  = £(l - e"kt)/be                          (4.5)
              L.   K.
where     C^  = concentration under the field
          R   = rate of pesticide loading

                             237

-------
          k   = decay rate
          t   = time of loading (4 months)
          b   = aquifer depth, and
          e   = porosity

The  concentration at the well can then be calculated using the
following relationship:
                      _kx
             C  = C- e  v                                (4.6)
              w    f
where     C   = concentration at well
           w
          Cf  = concentration under field
          k   = decay rate
          x   = distance to well from field
          v   = average velocity of water from field to well

Table  4.14  compares  those plug flow  concentrations  to  the
concentrations  determined  in the computer  simulations.   The
degree  of  agreement relates directly to how  much  dispersion
occurs  and dilutional differences between the single well used
in  CFEST and a line of wells assumed for the plug flow  model.
The  worst  cases  for  the Floridan  Aquifer  show  remarkable
agreement.   The  average cases for the Floridan  Aquifer  show
fairly  good  agreement.   The worst agreement is  between  the
cases for the surficial aquifers.

This   technique  of  calculating  concentrations  for   single
aquifers  with plug flow is useful for determining upper limits
on  the amount of pesticide contamination without the use of  a
sophisticated computer solute transport model.

Several  general trends can be noted in the scenarios simulated
with  the  computer.  The decay scenarios versus the  scenarios

                             238

-------
       TABLE 4.14  COMPARISON OF PEAK CONCENTRATIONS DETERMINED BY A SIMPLE PLUG
                   FLOW MODEL AND BY COMPUTER SIMULATION
                                           Predicted Well Water Concentrations
                                                          (ppb)
     Floridan worst case (deep well)
       No decay
       Decay
         300 ft well
         1000 ft well
                                           Plug Flow
                                         4.69

                                      1.05-1.11
                                       .35-.37
                                                      Computer Simulation
                      4.2-4.3

                      .91-.93
                      .33-.34
to
u>
vo
Floridan average cases (deep well)
  No decay
  Decay
    300 ft well
    1000 ft well
   9.36

0.13-0.28
  .000072
   3.3-8.1

 0.041-0.047
.000039-.00005
     Surficial worst case
       No decay
       Decay
         300 ft well
         1000 ft well
                                      62.5-187.5

                                      4.75-9.17
                                       .02-.74
                     28.1-94.7

                      1.4-1.5
                     .013-.015

-------
with  no decay have the greatest differences, the relative peak
concentrations   dropping  several  orders  of  magnitude  with
decay.   The  well parameters also show general trends  in  the
pumping  rates,  the  depth and the distance  from  the  source
area.   These  observations are enlarged upon in the  following
sections.

4.2.5.2  Pumping Rates—

The  results  from the simulations show that the pumping  rates
do  not  affect the peak relative concentration or the  general
shape  of  the  concentration profile.  In the  Floridan  worst
cases  the  greatest difference in peak concentrations  is  3%.
The  surficial  worst cases with no decay show  differences  as
much  as  50% between the high and low pumping rates, but  even
in  this  case, the trends are very similar.  The results  show
that  for  all the aquifer geometries, when there is no  decay,
the  scenarios  using  the lower pumping rate have  the  higher
peak  concentration.  Conversely, in the cases with decay,  the
scenarios  with  the higher pumping rate have the highest  peak
concentration.   With  no  decay, dilution  becomes  important.
The  wells  pumping at a higher rate bring more  "clean"  water
than  the  wells with lower rates, and consequently have  lower
concentrations.   This  effect  is especially apparent  in  the
worst  cases for the surficial aquifer, where the pumping  rate
dominates  the flow field as was shown in Figure 4.31.  In  the
worst  cases for the Floridan Aquifer the regional gradient  is
not affected very much by the well.

In  the  scenarios  with decay, the wells with  higher  pumping
rates  have  higher concentrations.  In those cases the  higher
pumping  rates produce higher ground-water velocities near  the
wells.    Higher   velocities  mean  short  travel  times   and
therefore less time for decay to occur.
                              240

-------
4.2.5.3  Well Depths—

Two  different well depths are simulated in the worst cases and
average  cases for the Floridan Aquifer.  In the Floridan worst
cases  the  concentrations are higher in the shallow well  than
the  deep  well, when the well is both 91 m and 300 m from  the
field.   In  these scenarios, the ground-water  velocities  are
high  enough  that not much downward dispersion  occurs  before
the  contaminant  reaches the well.  In the average  cases  for
the  Floridan Aquifer the concentrations are about equal in the
deep  well  and  shallow well when the well is 91  m  from  the
field.   When  the well is 300 m from the field, the deep  well
shows  higher  concentrations  than the shallow well.   In  the
average  cases  the ground-water velocities are slower than  in
the  worst  case.   This  means that there  is  more  time  for
downward  dispersion  to  occur.   At  91  m  from  the  field,
dispersion  has spread the contaminant in the three-dimensional
simulation  of  the shallow well, so that the concentration  is
about  equal  to  the  concentration in  the  deep  well.   The
difference   at  300  m  can  only  be  accounted  for  by  the
differences  in  aquifer  depths used in the  simulation.   The
shallow  well  is modeled as a partially penetrating well in  a
three-dimensional  aquifer  183  m  thick.  The  deep  well  is
modeled  in  a two-dimensional simulation of a 107  m  aquifer.
The  concentration in the deeper well is greater because it has
less aquifer depth in which to disperse.

4.2.5.4  Well Distances—

Two  different well distances are simulated with each scenario,
one  at 91 m from the source area and the other 300 m from  the
source  area.   As  might be expected, the  concentrations  are
higher  in  wells 91 m away.  In the cases with no  decay,  the
concentrations  at the 300 m well are smaller because there  is
more  time for dispersion.   The Floridan worst cases only show

                             241

-------
a  difference  of  about  20% between the 91 m and  the  300  m
cases.   The average Floridan and surficial worst case  differs
by  a  much greater amount (as much as 70%) due to the  greater
amount  of  time for dispersion.  In the cases with decay,  the
relative  concentrations  are  at least an order  of  magnitude
smaller  at  the 300 m well than at the 91 m well.  Again,  the
greater  amount  of time required to travel to the 300  m  well
allows more time for decay to occur.

4.2.5.5  Source Surrounding the Well—

In  the  production  runs,  the well  was  simulated  as  being
down-gradient  from  the edge of the grove.   Sensitivity  runs
were  performed  to determine the effect on peak  concentration
at  the  well  for  the case where the well  is  surrounded  by
citrus groves.

Two  sensitivity runs were made using the Surficial worst  case
(shallow  well,  200 gpm pumping rate, decay) and the  Floridan
average  case  (deep well, 1,000 gpm pumping rate,  no  decay).
These  cases  were chosen because the hydraulic  conductivities
are  low enough that pumping causes significant drawdown at the
well,  thereby  reversing the regional gradient.   Figure  4.40
compares  the  potential surfaces for two scenarios, one  where
the   regional   gradient  dominates,  the  other   where   the
well-induced  gradient  dominates the regional  gradient.   The
reversed  gradient  allowed the well to pull  contaminant  from
all  sides, not just from the up-gradient grove.  In both cases
groves  were assumed to surround the well at a distance of 91 m
(300 ft).

The  results  of  these  simulations (Figure  4.41)  show  that
aldicarb  is  pulled  from the down-gradient field as  well  as
from   the   up-gradient   field,  and  therefore,   the   peak
concentration  at  the well increases significantly.   For  the

                             242

-------
                            KH.E
                             1000
                             I
                              2000
                              j
    350.0
349.6
349.2   349.0
                               348.6   s* 348.2

                                PUMPING WELL


a.  Worst case for Floridan  aquifer.
                                           SCflLE
                                           200     400
                                           1	I
                                PUMPING WELL'


        b.  Worst  case for  surficial  aquifer.
Figure 4. 40  Comparison of  the two  extreme hydraulic

              potential distributions.
                             243

-------
3.
d

M
3.
Q

S
Z o.
2°

ff*
§"•*
_.
z
o
£*
(-•
tr
_i
"i
d
it
8.
d
t
d
o.
d
SOURCE SURROUNDING HELL „
\ .-'" ^"
.'' r»
z 2-
2
,' 1-0
,' or s
£ 2'
* »^
**" -i-
/'* ^-"""" ^ §_
' -^"^ £2 d
'' / S JS
/ X *—• Q
/ "*^*^ ^ °"

«s^*^ FAD(0)W(300. 50-350. 1000) ^
x i
/ X
x'>^ ^-

-•>fti> 	 , 	 , 	 , d


SOURCE SURROUNDING HELL
f '\ ,*>
, 1 / \ / \
/y / \ / \
/ \ / \ / \
/
/ %^*» '' s%** * "^»
^ »»•• *»»/ *»»«
/
i
;
*
f
1
1
t
1
r' SUD(29.7)V(300, 10-35.200)
A A7 A
/ \ ••' '•• / \
/ \ / \ /
1 \ / \ / \
. •-- 	 / . v-~,. /
0.0    180.0   360.0    540.0    720.0    900.0    1080.0
                TIME IN DflYS
O.C    18C.C    350.0    54C.C    720.C   90C.O    108C.C
                TIM.5 IK DflYS
  a.  Floridan average  case without
      decay.
b.   Surficial worst case  with decay.
          Figure 4.41   Effect of Aldicarb  source  area surrounding  the well.

-------
surficial  worst  case,  the relative peak increases  nearly  a
factor   of  4  over  the  case  where  the  source  was   just
up-gradient  from the well.  In the Floridan average case,  the
peak  is  not  reached in 3 years, however,  the  concentration
after  3  years  increases nearly a factor of 2 over  the  case
where  the  source  was  just up-gradient.  The  time  to  peak
remains the same in both cases.

The   results  show  that  peak  concentrations  can   increase
significantly  if the well is surrounded by the source area and
the  drawdown  at the well is large enough to  locally  reverse
the regional gradient.

4.2.5.6  Time to Reach a Peak Concentration—

For  some  of  the scenarios without  decay  (Floridan  average
case,  shallow  and  deep well; Surficial worst  case,  shallow
well)  a  peak was not reached at the pumping well  during  the
three-year   simulation   period.    In  order   to   gain   an
understanding  of  the magnitude of the peak and time  to  peak
for  these cases, the Floridan Average Case (shallow well,  200
gpm  pumping  rate, no decay, 300 ft from source to  well)  was
simulated  for  20  years rather than the 3 years used  in  all
other cases.

The  results  of this case are shown in Figure 4.42.  The  peak
concentration  at  the well plateaus after 13 years at a  value
of  about 17 ug/L.  This peak concentration is just over  twice
the  concentration  at  the  well after  3  years.   This  same
relationship  (i.e.,  actual peak concentration of about  twice
the  model  predicted  value, and a time to peak of  15  to  20
years)  can  probably  be applied to all the  Floridan  average
cases  where  the  peak  is  not  reached  in  the  three  year
simulation period.
                             245

-------
cn
              20-1
          D)
              15H
          CE
          uu
          o
              1CH
          8   5-,
          LLl
          Q_
                                            FAO(0)U(300.1»-40.200)
5          10         15

     TIME,   Yrs
                                                             20
          Figure 4.42  20-year  simulation of average   Floridan case,

-------
Model  runs  were not performed to determine the time  to  peak
for  the  Surficial  worst cases that do not reach  a  peak  in
three years.
4.2.6  Sensitivity Analysis and Verification
Several   sensitivity   runs   were  made  to   determine   the
sensitivity  of  the model results to assumptions made  in  the
modeling  concerning time step and input concentration and  the
validity  of  the assumption that the flow can be simulated  as
steady  state.   The sensitivity runs and steady  state  versus
transient comparisons are discussed in this section.

4.2.6.1  Time Step—

When  simulating  contaminant transport with a model that  uses
the  convective  dispersion  equation (such as  CPEST),  it  is
important   to  select  the  model  time  step  such  that  the
contaminant  travels  across  one  element  (i.e.,  the  finite
element  used to represent the model region) in one time  step.
If  this  condition  fails,  model results can  change  due  to
numerical  dispersion,  or by completely passing an element  in
one  time  step  and not using the hydraulic  conductivity  and
gradient  (i.e.,  ground-water velocity) associated  with  that
element.

The  time  steps  for the model runs made in  this  study  were
calculated   based  on the regional gradients assigned to  each
aquifer  case.   The  problem with this approach was  that  the
gradient  changes near the well as a result of the pumping, and
therefore,   the  time step near the well would not be  correct.
To  determine  if  the time step selected  could  significantly
alter  model  results, a series of sensitivity runs  were  made

                             247

-------
with a range of time  steps  for two different aquifer cases.

Six  sensitivity  runs were made with the Floridan  worst  case
(deep  well, 1,0000 gpm pumping, no decay) where the time  step
ranged   from 5 days to 60 days.  The optimum time step for this
case,  based on the regional gradient, was 10 days.  All  other
parameters in these simulations were identical.

The  results of these runs, as shown in Figure 4.43, show  that
the  peak  concentration  is not very sensitive to  changes  in
time  step.   The  peak concentrations at the well  changed  by
only  about  3%  and  6%, when the time  step  was  halved  and
doubled,  respectively.   The  trend  is  such  that  the  peak
concentration decreases as  the time step increases.

The  Floridan  worst  case  had less than 0.5  m  (1.5  ft)  of
drawdown  at  the well, therefore, the change in gradient,  and
the  resulting  change  in  travel time, was  quite  small.   In
order   to  better  test  the  impact  of  time  step  on  peak
concentration,  a  series of 5 sensitivity runs were  made  for
the  Surficial  worst case  (shallow well, 200 gpm  pumping,  no
decay)  where  the drawdown at the well was about 3 m (10  ft).
The  optimum time step for  this case was 30 days and time steps
of  5  to  60 days were tested in the  sensitivity  runs.   All
other parameters in these simulations were identical.

The  results  of these simulations, also in Figure  4.43,  show
that,  as  in the earlier case, the peak concentration  is  not
very  sensitive  to changes in the model time step.   The  peak
concentration   at   the  well  decreased  as  the  time   step
increased, but the rate of change was small.

Both  the  Floridan  worst case and the  surficial  worst  case
showed  a low degree of sensitivity to change in time step, and
the  percentage  change in peak concentration at the  well  was

                            248

-------
to
4*
V£>
              D)
              13
                 4-
              <
              cc
LU
CJ
Z

S
              UJ
              Q.
                 3-
                                                 Legend
                                                 A FLOHIDAN

                                                 X SURFICIAL
                         10     20     30    40     50
                              TIME STEP,  Days
                                             60
              Figure 4.43
             Relationship of model time step to peak concentrations
             for the Floridan and surficial worst cases.

-------
about  the same for both cases.   Part of the reason that  time
step  is  not a significant fact in these simulations  is  that
the  hydraulic  conductivity  is uniform within  each  aquifer.
Even   if  the  distance  traveled  by  the  ground  water  and
contaminant  is greater than one element spacing, it will still
be  using  the  same hydraulic conductivity  to  calculate  the
velocity.   Therefore,  running  all cases with a  single  time
step  based on the regional gradient, and using a uniform  grid
spacing  over the entire aquifer, does not impact the  validity
of the model results.

4.2.6.2  Input Concentration—

A  sensitivity  run was made to show that a  direct  one-to-one
relationship  exists  between concentration at the  source  and
peak   concentration   at  the  well.   This  relationship   is
important  because all the saturated model runs were made  with
a   unit   loading  at  the  field,  and  the  predicted   peak
concentrations  at  the  well  must  be  scaled  to  the  field
loadings as predicted by the unsaturated zone analysis.

A  single  run  was made using the Floridan  worst  case  (deep
well,   1,000   gpm   pumping  rate,  no   decay)   where   the
concentration  at  the field was doubled every time  step,  and
the  peak  concentration  at the well, were exactly  twice  the
values  obtained  from  the original Floridan worst  case  run.
Because  this direct relationship holds, peak concentrations at
the   well  can  be  scaled  directly  to  the   concentrations
predicted by the unsaturated zone model.
4.2.6.3  Transient Versus Steady State Comparison—

4.2.6.3.1   Ground-water  Potential—For the purposes  of  this
study  recharge was assumed to be evenly distributed throughout
the   year,   and  therefore,   the  ground-water  potential  was
modeled   as   steady  state.    In  actuality,    rainfall   and

                            250

-------
irrigation  occur as events and are seasonal, which results  in
a  nonuniform  distribution  of water to  the  ground  surface.
These  pulses  of  water at the surface are  smoothed  as  they
travel  through  the  unsaturated zone, resulting in  a  fairly
uniform  distribution of recharge to the water table,  however,
occasional  pulses  do  occur.   If it can be  shown  that  the
impact  of  these  pulses is short lived (i.e.,  the  resulting
ground-water  mounds  rapidly  decay)  then  the  steady  state
assumption should be valid.

Two  model  simulations  were made using the  Floridan  average
case   (deep  well,  1,000  gpm  pumping  rate,  no  decay)  to
determine  the time required for a ground-water mound to decay.
In  both simulations the hydraulic conductivity was reduced  to
152 m/day (500 ft/day) to increase the impact of the recharge.

The  first  simulation  consisted of running the  steady  state
model  with  33  cm (13 in) of recharge  distributed  uniformly
throughout  the  year.   The  second  simulation  consisted  of
running  the  transient  model with 33 cm  of  recharge  evenly
distributed  throughout the year,  in addition to a 15 cm (6 in)
pulse  on  the  first  day.  The 15 cm  pulse  is  the  maximum
probable  daily pulse of water as predicted by the  unsaturated
flow  model  over a 14 yr period.   The transient model was  run
until  the  predicted  potential surface was identical  to  the
steady state surface.

The  results of the transient simulation showed that it took 37
days  for the potential surface to exactly equilibrate with the
surface  predicted  by the steady state model.  After  10  days
the  transient surface lost about 80% of its initial rise,   and
95% was lost after 20 days.

The  15 cm pulse used in this simulation is an extreme  maximum
that  would  be  expected on a given day.    The  typical  daily

                            251

-------
pulse,  predicted  by the unsaturated model, is  less  than  one
tenth  this  amount,  and the pulse rarely  exceeds  half  this
amount.   Given these facts, it is probably safe to assume that
the  impact  of  most pulses will be negligible  within  a  few
days.

Since  the  unsaturated  zone has the effect of  smoothing  the
recharge  distribution, and most pulses of recharge equilibrate
within  a  few  days,  the steady state  assumption  should  be
valid.
                         *
4.2.6.3.2	Contaminant   Transport—Simulation   with    the
unsaturated  model showed that about 90% of the aldicarb enters
the  ground  water within the first 4 months of each year,  and
that  the highest annual loading rate is about 1 x 10   gm/cm  .
As  a  result of these findings, this annual loading  rate  was
used  in  all  saturated  flow model simulations,  and  it  was
distributed uniformly over four months of each year.

To  test the validity of this assumption, a simulation was made
where  the  actual  annual loading rates as  predicted  by  the
unsaturated  model were simulated in the saturated flow  model.
A  two-year simulation was made using the 1970 and 1971  output
from  the  unsaturated  flow model in the Floridan  worst  case
(deep  well, 1,000 gpm pumping rate, no decay).  The model  was
run  with  a 5 day time step, therefore,  contaminant  loadings
were  simulated  as  the  total  loading  over  each  five  day
interval.   These  results were compared to a model  run  where
the  1970  and  1971  annual loading rates  were  simulated  as
uniformly  entering  the ground water over four months of  each
year.

The  results of these simulations are shown in Table 4.15.  For
both  years,  the results (both relative peak concentration  at
the  well and time to peak) for the average loading  simulation

                            252

-------
  TABLE  4.15  RESULTS OF  AVF.RAGE VERSUS ACTUAL
               ALDICARB MASS LOADING SIMULATIONS
        Relative Peak  Concentration     Time to Peak (days)
            AverageActualAverage   Actual
Year        Loading      Loading         Loading   Loading

1970      9.0 x 10"3   9.8 x 10"3         150       135

1971      8.9 x 10"3   5.4 x 10"3         150       220
                           253

-------
are   close  to  those  for  the  actual  loading   simulation,
indicating that the averaging assumption should be valid.
4.3  COMBINED RESULTS OF UNSATURATED AND SATURATED ZONE
     MODELING
4.3.1  Physical Overlap of Saturated and Unsaturated Zones
The   results  from  the  unsaturated  zone  modeling  and  the
saturated  zone  modeling  efforts were  combined  to  estimate
aldicarb   concentrations   in  drinking  water   wells.    The
unsaturated  zone scenarios were reduced to two physical areas:
1)   areas in which entisols and ultisols are prevalent, and 2)
areas  in which spodosols and alfisols are prevalent.  For  the
entisols  and ultisols both a thin and a thick unsaturated zone
were  considered  but  the physical area is  the  same  (Figure
4.44).   In  the saturated zone the five aquifer  systems  were
combined   into  three  physical  aquifer  systems  that   were
modeled:    1)   the  unconfined  Floridan  Aquifer,   2)   the
unconfined  surficial aquifer and 3) the multi-aquifer  system.
These  are  shown in Figure 4.44.  Figure 4.45 shows the  areas
of  overlap  of the unsaturated and saturated  zone  scenarios.
The  entisols  and  ultisols overlap all three of  the  aquifer
systems.    The   spodosols  and  alfisols  overlap  with   the
surficial  aquifer  and the multiaquifer systems but  not  with
the  unconfined  Floridan  Aquifer.   The results  of  the  two
modeling  studies  were  combined in all  cases  where  overlap
occured.
                             254

-------
                                                   Atlantic Octal
         ENTISOLS AND ULTISOLS
         SPODOSOLS AND ALFISOLS
                                                   Atlantic Ocean
                LEGEND





        FLORIDIAN AQUIFER - UNCONFINED





        MULTI - AQUIFER  SYSTEM





        SURFICIAL AQUIFER - UNCONFINED
                                             ./» *
Figure 4.44  Unsaturated and  saturated zone groupings.
                           255

-------
                ALABAMA
                        Gulf of Mexico

                                                     Jacksonville
                                                         Atlantic Ocean
                       LEGEND


               ENTISOLS AND ULTISOLS


               FLORIDIAN AQUIFER - UNCONFINED


               MULTI - AQUIFER SYSTEM
               SURFICIAL AQUIFER - UNCONFINED
Figure 4.45  Overlap of unsaturated  and saturated model
               areas.
                                 256

-------
4.3.2  Method of Combining the Modeling Results
Frequency   distributions   of  pesticide  loading   from   the
unsaturated  zone and the relative pesticide concentrations  in
the  well  water were developed from model results.   Frequency
distributions   for  relative  well-water  concentrations   are
located  in  Appendix  E.  A statistical approach was  used  so
that  the  risks  of  exposure could  be  accurately  assessed.
Because  the  extreme values are of greatest interest  in  this
study,  the  ten percent exceedance value was used  along  with
the  50  percent  exceedence value for combining  the  modeling
results.   The unsaturated zone values were shown previously in
Table  4.5.   The 10 and 50 percent exceedance values  for  the
saturated zone scenarios are shown in Appendix E.

The  well  concentrations were calculated using  the  following
relationship:

           c  =  f  •  c   n                                 (4'7)
            w        rel
where     C   = well water concentration
          C r 1= relative well water concentration
          f   = a conversion factor equal to the
                input concentration x unsaturated load
                divided by the unit load

The  concentration in the well is determined from the  relative
concentration  and a conversion factor which is the product  of
the  input  concentration and the unsaturated load  divided  by
the  unit  load  "used in the saturated zone  simulations.   The
input  concentrations  used in the saturated  zone  simulations
were  shown  in  Table 4. 8  .  The unit load used  in  saturated
zone simulations was always 1 kg/ha.  Using  equation 4.7, the
                              257

-------
resulting  well water concentration for any pesticide load  can
be calculated.
4.3.3  Results
The   complete   set   of  results  showing  the   well   water
concentration  at  the 10 and 50 percent exceedance values  for
all  the scenarios are shown in Table 4.16.  These results  are
summarized  in Table 4.17 which reports the highest  calculated
concentration  for  each catagory shown.  In general  the  well
water  concentrations are very low.  The highest  concentration
is  6.5  parts per billion (ppb) in a scenario  without  decay.
This  is  just at the detection limit of about 5 to 6  ppb  for
total  toxic residues (Rao, 1984, personal communication).  The
highest  concentrations  for  scenarios with decay are  in  the
order of 10"  ppb.

The  well  water concentrations show three general trends.   As
shown  earlier,  in  the  discussion of  the  unsaturated  zone
results,  the highest loads are leached from the thin  ultisols
and  entisols, followed by the thick ultisols and entisols  and
the  alfisols and spodosols.  In the saturated zone, the  worst
cases  of the surficial aquifer generally have the highest well
water   concentrations  closely  followed  by  the   unconfined
Floridan  worst  cases  and  finally the  worst  cases  of  the
two-aquifer  system  and average Floridan cases.   Within  each
set  of saturated zone scenarios there is a third trend that is
consistent  for  all the aquifers.  The highest  concentrations
result  from  the no decay simulation with the well 91  m  (300
ft)  from the source area, followed by no decay simulated  with
a  well at 300 m  (1000 ft).  Next are the decay scenarios  with
the  well at 91 m.  The smallest concentrations result from the
decay simulations where the well is 300 m from the source.

                              258

-------
          TABLE  4.16  WELL WATER CONCENTRATIONS  (PPB)   FOR  COMBINED  UNSATURATED  AND  SATURATED
                          SCENARIOS  AT  50  AND  10  PERCENT  EXCEEDANCE PROBABILITIES  OF  EACH.
to

-------
          TABLE  4.16  (CONT.)
to
CTl
O
                  Saturated Zone
Surficial Aquifer, worst case
  91 meters to well
    High pumping rate, no decay 50%
    High pumping rate, no decay 10%
    Low pumping rate, no decay 50%
    Low pumping rate, no decay 10%
    Low pumping rate, decay 50%
    Low pumping rate, decay 10%

  300 meters to well
    High pumping rate, no decay 50%
    High pumping rate, no decay 10%
    Low pumping rate, no decay 50%
    Low pumping rate, no decay 10%
    Low pumping rate, decay 50%
    Low pumping rate, decay 10%
          Two-aquifer System, worst case
            91 meters to well
              High pumping rate, no decay 50%
              High pumping rate, no decay 10%
              Low pumping rate, no decay 50%
              Low pumping rate, no decay 10%
              Low pumping rate, decay 50%
              Low pumping rate, decay 10%
                                                Alfisols and Spodosols
                                                   50%         10%
7.98x10-2
1.17x10-2
2.11x10-2
2.7x10-2
1.69x10-4
5.27x10-4
4.21x10-3
9.42x10-3
7.47x10-3
1.64x10-2
2.92x10-6
5.20x10-6
1.65x10-4
3.95x10-4
2.68x10-4
4.91x10-4
1.56x10-5
6.37x10-5
4.39x10-2
6.41x10-2
1.16x10-1
1.49x10-1
9.30x10-4
2.90x10-3
2.32x10-2
5.18x10-2
4.11x10-2
9.04x10-2
1.61x10-5
2.86x10-5
9.07x10-4
2.17x10-3
1.48x10-3
2.70x10-3
8.57x10-5
3.51x10-4
                                                                           Unsaturated Zone

                                                                    Thick Ultisols  and Entisols

                                                                         50%              10%
                                                                                8.38x10-3
                                                                                1.22x10-2
                                                                                2.21x10-2
                                                                                2.85x10-2
                                                                                1.78x10-4
                                                                                5.53x10-4
4.42x10-3
9.89x10-3
7.84x10-3
1.73x10-2
3.07x10-6
5.46x10-6
                                                                      1.73x10-4
                                                                      4.15x10-4
                                                                      2.82x10-4
                                                                      5.16x10-4
                                                                      1.64x10-5
                                                                      6.69x10-5
                6.36x10-2
                9.32x10-2
                1.69x10-1
                2.17x10-1
                1.35x10-3
                4.21x10-3
3.37x10-2
7.53x10-2
5.98x10-2
1.32x10-1
2.34x10-5
4.16x10-5
                1.32x10-3
                3.16x10-3
                2.15x10-3
                3.93x10-3
                1.25x10-4
                5.10x10-1
                                 Thin  Ultisols  and Entisols
                                      50%             10%
                   2.59x10-1
                   3.79x10-1
                   6.85x10-1
                   8.82x10-1
                   5.5x10-3
                   1.17x10-2
1.92
2.8
5.06
6.51
4.06x10-2
1.26x10-1
1.37x10-1
3.06x10-1
2.43x10-1
5.34x10-1
9.49x10-5
1.69x10-4
1.01
2.26
1.79
3.95
7.01x10-4
1.25x10-3
5.36x10-3
1.28x10-2
8.72x10-3
1.60x10-2
5.06x10-4
2.07x10-3
3.96x10-2
9.49x10-2
6.44x10-2
1.18x10-1
3.74x10-3
1.53x10-2

-------
TABLE 4.17  HIGHEST CALCULATED ALDICARB CONCENTRATIONS  (in ppb)
            IN THE GIVEN COMBINED UNSATURATED/SATURATED CATAGORIES

Floridan
Worst Cases
Floridan
Average
Surf icial
Worst Cases
Two-
Aquifer
Worst
Cases
300
ft
1000
ft
300
ft
1000
ft
300
ft
1000
ft
300
ft
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
No
Decay
Decay
Flatwoods
Soils
No
Overlap
No
Overlap
l.Sxlo'1
2.9xlO~3
9.0xlO~2
2.7x10-*
2.7xlO~3
3.5xlO~*
Thick
Ridge Soils
2.3xlO~2
4xlO~3
1.8xlO~2
9.8xlO~*
2.2xlO~2
1.3x10""
8.8xlO~3
1.4xlO"7
2.2xlO-1
4. 2xlO~ 3
1.3X10"1
4.2xlO~5
3.9xlO~3
5.1xlO~"
Thin
Ridge Soils
6.9x10'*
1.2X10"1
5.5x10"'
2.9xlO~2
6.5xlO~a
3.9xlO~3
2.6xlO-1
4. xlO~6
6.5
l.SxlO"1
3.9
1.3xlO~3
1.2X10'1
1.5xlO~2
                               261

-------
Surficial Aquifer
Combined  with  the thin entisols and ultisols,  the  surficial
aquifer    worst   cases   generally   produce   the    highest
concentrations.   The highest value is for worst case hydraulic
parameters,  no  pesticide decay, with a shallow well at  91  m
distance  from the source area, pumping at a low rate having  a
well  water concentration of 6.5 parts per billion (ppb).   With
decay   the   concentrations  drop  one  to  three  orders   of
magnitude.   When  the  well  is 300 m  from  the  source,  the
surficial   cases   without   decay  still   show   significant
contamination,  but when decay is simulated the  concentrations
are  in the order 10   ppb or less.  The other unsaturated zone
scenarios  show  the same trends as the thin ridge  soils  when
combined  with  the  surficial aquifer worst  cases  with  even
lower  concentrations.   The highest concentrations are 2.17  x
  -1                  -1
10    and  1.49  x  10    for the thick  ridge  soils  and  the
flatwoods soils, respectively.

Floridan Aquifer Worst Cases
Well  water concentrations in the Floridan worst cases are  not
as  high  as  the surficial worst cases except  when  decay  is
modeled   with  the  drinking  water  well  at  300  m.    When
evaluating  these  results  with no decay, it is  necessary  to
recall  that  these  concentrations could be  much  higher  for
wider   source   areas.   The  highest  concentration  in   the
scenarios  without decay is 6.9 x 10   ppb with the well at  91
m   (300  ft) and 5.5 x 10~1ppb with  the well at 300  m  (1000
ft).   The  scenarios  simulated with decay are  generally  one
order of magnitude less than the same scenario without decay.

The Floridan Aquifer Average Cases
The  well  water concentrations in the Floridan  average  cases
vary  considerably  between  scenarios modeled with  decay  and
without  decay.   The concentrations calculated  for  scenarios

                            262

-------
with  the  well 91 m (300 ft) away from the source area and  no
decay  are  as high as 0.65 ppb.  In the cases with  decay  the
concentrations  drop over two orders of magnitude when the well
is  91  m from the source and four to five orders of  magnitude
when the well is 300 m from source.

Two-Aquifer System
The  scenarios modeled with the two-aquifer system have some of
the  lowest  concentrations.   Recall that  the  concentrations
were  suspected of being higher than they should be due to  the
hydraulics  of the system.  Even so the concentrations are low.
The  highest  value  is  0.12 ppb.  With  decay  simulated  the
concentrations drop over an order of magnitude.
                             263

-------
                          REFERENCES
Anderson, C.A. 1981.  Fertility and Management for Citrus on
    Entisols.  In:  Proceedings of Symposium on Fertility
    Management of Florida Soils.  Soil and Crop Science
    Society of Florida, Vol 40.

Anderson, W. and E.P. Simonds.  1982.  Hydrology of Hamilton
    Lakes and Vicinity, Polk County, Florida.  U.S. Geological
    Survey Water-Resources Investigations Open-File Report
    82-630.

Bearden, H.W.  1972.  Water Available in Canals and Shallow
    Sediments in St. Lucie County, Florida.  Florida Bureau of
    Geology Report of Investigation No. 62.

Biggs, R.H. and P.G. Webb.  1983.  Fate of Aldicarb in Plants
    and the Nature of Citrus Responses.  In:  Aldicarb
    Research Task Force Report.  IFAS, University of Florida,
    Gainesville.

Brady, N.C.  1974.  The Nature and Properties of Soils.
    McMillan Publishing Co., Inc.  8th Edition.

Bromilow, R.H., R.J. Baker, M.A.H. Freeman and K. Gorog.
    1980.  The Degradation of Aldicarb and Oxamyl in Soil.
    Pesticide Sci. 11: 371-378.
                              264

-------
Bull, D.L., R.A. Stokes, J.R. Coppedge and R.L. Ridgway.
    1970.  Further Studies of the Fate of Aldicarb in Soil.
    J. of Economic Entomology.  63(4):1283-1289.

Buono, A., R.M. Spechler, G.L. Barr and R.M. Wolsansky.  1979.
    Generalized Thickness of the Confining Bed Overlying the
    Floridan Aquifer, Southwest Florida Water Management
    District.  U.S. Geological Survey Water-Resources
    Investigations Open-File Report 79-1171.

Bush, P.W.  1982.  Predevelopment Flow in the Tertiary
    Limestone Aquifer, Southeastern United States; A Regional
    Analysis from Digital Modeling.  U.S. Geological Survey
    Water-Resouces Investigation 82-905.

Caldwell, R.E. and R.W. Johnson.  1982.  General Soil Map of
    Florida.  United States Dept. of Agriculture, Soil
    Conservation Service.

Calhoun, F.G., V.W. Carlisle, R.E. Caldwell, L.W. Zelazny,
    L.C. Hammond and H.L. Breland.  1974.  Characterization
    Data for Selected Florida Soils.  Soil Science Research
    Report No. 74-1, University of Florida, IFAS.

Calvert, D.V., R.C.J. Koo and H.W. Ford.  1967.  Flood
    Irrigation Studies with Citrus.  Proc. Fla. State Hort.
    Soc.  80:79-85

Carlisle, V.W., R.E. Caldwell, F. Sodek, III, L.C. Hammond,
    F.G., Calhoun, M.A. Granger and H.C. Breland.  1978.
    Characterization Data for Florida Soils.  Soil Science
    Research Report No. 78-1, University of Florida, IFAS.
                              265

-------
Carlisle, V.W., C.T. Hallmark, F. Sodek III, R.E. Caldwell,
    L.C. Hammon and V.E. Berkheiser.  1981.  Characterization
    Data for Selected Florida Soils.  Soil Science Research
    Report No. 81-1, University of Florida, IFAS.

Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean and P.P.
    Jowise.  1984.  User's Manual for the Pesticide Root Zone
    Model  (PRZM) Release I.  U.S. Environmental Protection
    Agency, Athens, GA.

Causey, L.V. and G.W. Leve.  1976.  Thickness of the Potable-
    Water Zone in the Floridan Aquifer.  Florida Bureau of
    Geology Map Series 74.

Cherry, R.N., J.W. Stewart and J.A. Mann.  1970.  General
    Hydrogeology of the Middle Gulf Area, Florida.  Florida
    Bureau of Geology Report of Investigation, No. 56.

Chow, V.T.  1964.  Handbook of Applied Hydrology.  McGraw-Hill
    Book Co.  New York, NY.

Cohen, S.Z., S.M. Creeger, R.F. Carsel and C.G. Enfield.
    1983.  Potential for Pesticide Contamination of Ground
    Water Resulting from Agricultural Uses.

Cooke, C.W.  1945.  Geology of Florida: Florida Geological
    Survey Bull. 29.

Grain, L.J., G.H. Hughes and L.J. Shell.   1975.  Water
    Resources of Indian River County, Florida.  Florida Bureau
    of Geology Report of Investigation, No. 80.
                              266

-------
Duerr, A.D. and J.T. Tromraer.  1982.  The Benchmark Farm
    Program:  A Method for  Estimating Water Use in Southwest
    Florida.  USGS Water-Resource Investigation 82-17.
    U.S. Geological Survey, Tallahassee, FL.

Fayard, L.D., D.M. Bonnet and R.G. Belles.  1983.  Water
    Resources Data, Florida, Water Year 1982, Vol. IB,
    Northeast Florida Groundwater.  U.S. Geological Survey
    Water-Data Report FL-82-1B.

Florida Citrus Mutual.  1984.  Lakeland, FL.

Florida Dept. of Agriculture and Consumer Services.   1984.
    Citrus Summary 1983.  Orlando, FL.

Gupta, S.K., C.T. Kincaid, P.R. Meyer, C.A. Newbill and C.R.
    Cole.  1982.  A Multi-Dimensional Finite Element  Code for
    the Analysis of Coupled Fluid, Energy and Solute  Transport
    (CFEST).  PNL-4260, Pacific Northwest Laboratory,
    Richland, WA.  Prepared for the U.S. Dept. of  Energy.

Hansen, J.L. and M.H. Spiegel.  1983.  Hydrolysis  Studies of
    Aldicarb, Aldicarb Sulfoxide and Aldicarb Sulfone.  Env.
    Tox. and Chem. 2:147-153.

Harrison,  D.S.  1984.  Citrus Water Use and
    Evapotranspiration  (ET) Requirements for Florida
    Flatwoods Soils.  Florida Cooperative Extension Service
    Extension Report  84-1.

Harrison,  D.S., J.F.  Gerbeer and R.E. Choate.   1974.
    Sprinkler Irrigation  for Cold  Protection.   Technical
    Circular 348.  Florida  Cooperative Extension  Service.
                              267

-------
Healy, H.G.  1974.  Water Wells in the Artesian and
    Nonartesian Aquifers of Florida, 1971-1972.  USGS, Bureau
    of Geology.  Information Circular No. 85.

Healy, H.G.  1975.  Piezometric Surface and Areas of Artesian
    Flow of the Floridan Aquifer in Florida.  Florida Bureau
    of Geology Map Series No. 4, second edition.

Healy, H.G.  1982.  Surficial and Intermediate Aquifers in
    Principal Aquifers in Florida, B.J. Franks, ed.  U.S.
    Geological Survey Water-Resources Investigations Open-
    File Report 82-255.

Hem, J.D.  1970.  Study and Interpretation of the Chemical
    Characteristics of Natural Water.  U.S. Geological Survey
    Water Supply Paper 1473.

Hornsby, A.G., P.S.C. Rao, W.B. Wheeler, P. Nkedi-Kizza and
    R.L. Jones.  1983.  In Press.  Fate of Aldicarb in Florida
    Citrus Soils  1.  Field and Laboratory Studies.
    Proceedings of the NWWA Symposium on Characterization and
    Monitoring of the Unsaturated  (Vadose) Zone.  D. Nielsen,
    ed.

Hough, A.,  I.J. Thomason and W.J. Farmer.   1975.  Behavior  of
    Aldicarb in Soil Relative to Control of Heterodera
    Schachtii.  J. of Nematology.  7(3):214-221.

Hughes, G.H.   1977.  Runoff  from Hydrologic Units in Florida.
    Florida Bureau of Geology Map  Series No. 81.

Hutchinson, C.B.  1978.  Appraisal of Shallow  Ground-Water
    Resources  and Management Alternatives  in the Upper Peace
    and Eastern Alafia River Basins, Florida.   U.S. Geological
    Survey, Water-Resources  Investigations, pp. 77-124.

                              268

-------
INTERA Environmental Consultants.  1980.  Mathematical
    Simulation of Aldicarb Behavior on Long Island:
    Unsaturated Flow and Ground-Water Transport.  Prepared for
    the Hayzard Evaluation Division, U.S. Environmental
    Protection Agency.

Jones, R.L., P.S.C. Rao and A.G. Hornsby.  1984.  In Press.
    Fate of Aldicarb in Florida Citrus Soil 2. Model
    Evaluation.  Proceedings of the NWWA Synpvision on
                                                              -j.
    Characterization and Citation Monitoring of the
    Unsaturated (Vadose) Zone.  D. Nielsen, Ed.

Klein, H., M.C. Schroeder, W.F. Lichtler.  1964.  Geology and
    Ground-Water Resources of Glades and Hendry Counties,
    Florida.  Florida Geological Survey Report of
    Investigation, No. 37.

Knochenmus, D.D. and G.H. Hughes.  1976.  Hydrology of Lake
    County, Florida.  U.S. Geological Survey Water-Resources
    Investigations, 76-72.

Kohler, M.A., T.J. Nordenson and D.R. Baker.  1959.
    Evaporation Maps for the United States.  U.S. Weather
    Bureau Tech. Paper 37.

Land, L.F., H.G. Rodis and J.J.  Schneider.  1973.  Appraisal
    of the Water Resources of Eastern Palm Beach County,
    Florida.  Florida Bureau of  Geology Report of
    Investigations, No. 67.

Leistra, M., T. Lexmond and J.H. Smelt.   1976.  Conversion and
    Leaching of Aldicarb in Soil Columns.  Pesticide Sci.
    7(5):  471.
                              269

-------
Lemley, A.T. and W.Z. Zhong.  1983.  Kinetics of Aqueous Base
    and Acid Hydrolysis of Aldicarb, Aldicarb Sulfoxide, and
    Aldicarb Sulfone.  Env. Tox and Chem. 2:147-153.

Leraley, A.T. and W.Z. Zhong.  1984.  Hydrolysis of Aldicarb,
    Aldicarb Sulfoxide and Aldicarb Sulfone at ppb Levels in
    Aqueous Media.  J. Ag. and Food Chem, July-August,  1984.

Lichtler, W.F.  1960.  Geology and Ground-Water Resources of
    Martin County, Florida.  Florida Geological Survey  Report
    of Investigation, No. 23.

Lichtler, W.F.  1972.  Appraisal of Water Resources in  the
    East Central Florida Region.  Florida Bureau of Geology
    Report of Investigation, No. 61.

Lyman, W.J. 1982.  Adsorption Coefficient for Soils and
    Sediments.  Chapter 4 In:  Handbook of Chemical Property
    Estimation Methods.  W.J. Lyman, W.F. Rheel and D.H.
    Rosenblatt, eds.  McGraw-Hill Book Co., New York, NY.

Merritt, M.L., F.W. Meyer, W.H. Sonntag and D.J. Fitzpatrick.
    1983.  Subsurface Storage of Freshwater in South  Florida:
    A Prospectus.  U.S. Geological Survey Water-Resources
    Investigation, 83-4214.

Myeyk, R.T., L.D. Fayard, W.L. Fletcher and J.K. Ogle.   1983.
    Water Resources Data:  Florida, Water Year 1982,  Vol. 3B.
    Southwest Florida Ground Water.  U.S. Geological  Survey
    Water-Data Report FL-82-3B.

Nicholls, P.H., R.H. Bromilow and T.M. Addiscott.   1982.
    Measured and Simulated Behavior of Fluometuron, Aldoxycarb
    and  Chloride Ion in a Fallow Structured Soil.   Pestic.
    Sci. 13: 475-483.

                              270

-------
Pacenka, S. and K.S. Porter.  1981.  Preliminary Regional
    Assessment of the Environmental Fate of the Potato
    Pesticide, Aldicarb, Eastern Long Island, New York.
    Cornell University, Center for Environmental Research.

Parker, G.G.,  G.E. Ferguson and S.K. Love.  1955.  Water
    Resources of Southeastern Florida, with Special
    Reference to the Geology and Ground Water of Miami Area:
    U.S. Geological Survey Water-Supply Paper 1255.

Phelps, G.G.  1978.  Chemical Quality of Water used for
    Municipal Supply in Florida, 1975.  U.S. Geological Survey
    Map Series, No. 82.

Porter, K.S., A.T. Lemley, H.B. Hughes and R.L. Jones.  1984.
    Developing Information on Aldicarb Levels in Long Island
    Groundwater.  Second International Conference on Ground-
    Water Quality Research, Proceedings.  March 1984.  Ada,
    OK.

Rao, P.S.C., L.T. Ou, K.S.V. Edvardsson, J.T. Thomas and W.B.
    Wheeler.  1984.  Degradation and Sorption of Aldicarb  and
    Metolachlor in Dougherty Plains Soils.  Progress Report
    dated June 15, 1984.  For the  U.S. Environmental
    Protection Agency, Athens, GA.

Ryder,  P.D.  1982.  Digital Model  of Predevelopment Flow in
    the Tertiary Limestone  (Floridan) Aquifer System in
    West-Central Florida.  U.S. Geological Survey Water-
    Resources Investigations 81-54.

Scott,  W.B.  1977.  Hydraulic Conductivities and Water Quality
    of  the  Shallow Aquifer, Palm Beach County, Florida.  U.S.
    Geological Survey Water-Resources Investigation 76-119.

                             271

-------
Sinclair, W.C.  1974.  Hydrogeologic Characteristics of the
    Surficial Aquifer in Northwest Hillsborough County,
    Florida.  Florida Bureau of Geology Information.  Circular
    No. 86.

Smajstrla, A.G., D.S. Harrison, C. Tai and D. Clapp.  1982.
    Water Budget of Crown Flood Irrigated Citrus.
    Proceedings of the Florida State  Horticultural Society,
    Vol. 95.

Smelt, J.H., A. Dekker, M. Leistra and N.W.H. Houx.  1983.
    Conversion of Four Carbamoyloximes in Soil Samples from
    Above and Below the Water Table.  Pesticide Science 14:
    173-181.

Smelt, J.H., M. Leistra, N.W.H. Houx and A. Dekker.  1978a.
    Conversion Rates of Aldicarb and Its Oxidation Products in
    Soils. I.  Aldicarb Sulfone.  Pesticide Sci. 9:279-285.

Smelt, J.H., M. Leistra, N.W.H. Houx and A. Dekker.  1978b.
    Conversion Rates of Aldicarb and Its Oxidation Products
    in Soils. II.  Aldicarb Sulfoxide.  Pesticide Sci.
    9:286-292.

Smelt, J.H., M. Leistra, N.W.H. Houx and A. Dekker.  1978c.
    Conversion Rates of Aldicarb and Its Oxidation Products in
    Soils. III.  Aldicarb.  Pesticide Sci. 9:293-300.

Stanley, J.M., C. Taylor, W.R. Sununerhill, Jr. and  L.J.
    Beaulieu.  1980.  Citrus Energy Survey - Use Estimates
    and Conservation.   Institute of Food and Agricultural
    Sciences.  University of Florida, Energy Report No. 2.
                             272

-------
Stewart, H.G.  1966.  Ground-Water Resources of Polk County.
    Florida Geological Survey, Report of Investigation, No.
    44.

Stewart, J.W.  1980.  Areas of Natural Recharge to the
    Floridan Aquifer in Florida.  Florida Bureau of Geology
    Map Series 98.

Stringfield, V.T.  1966.  Artesian Water in Tertiary Limestone
    in the Southeastern States.  U.S. Geological Survey
    Professional Paper 517.

Supak, J.R.  1972.  The Volatilization, Degradation,
    Adsorption and Desorption Characteristics of Aldicarb in
    Soils and Clays.  Ph.D Dissertation, Texas A&M University.

Supak, J.R., A.R. Swoboda and J.B. Dixon.  1977.
    Volatilization and Degradation Losses of Aldicarb  from
    Soils.  J. of Environ. Qual. 6(4):413-417.

Thibodeaux, L.J.  1979.  Chemodynamics.  John Wiley and Sons.
    New York, NY.

Tibbals, C.H.  1981.  Computer Simulation of the Steady-State
    Flow System of the Tertiary Limestone (Floridan) Aquifer
    System in East-Central Florida.   U.S. Geological Survey
    Water-Resources Investigations Open-File Report 81-681.

U.S.  Dept. of Commerce.  1968.  Climatic Atlas of the  United
    States.

U.S.  Dept. of Commerce.  1972.  Climate of the States,
    Florida:  Climatography of the United States, No.  60-8,
    National Oceanic  and Atmospheric  Adm. Environmental Data
    Service.

                              273

-------
U.S. Geological Survey.  1982.  Water Resources Data:
    Florida, Water Year 1981.  U.S. Geological Survey
    Water-Data Report, Vol. 1, 2B and 3B.

Viessman, Jr., W., J.W. Knapp, G.L. Lewis and T.E. Harbaugh.
    1977.  Introduction to Hydrology.  2nd edition.  Harper &
    Row, NY.

Visher, F.N. and G.H. Hughes.  1969.  The Difference Between
    Rainfall and Potential Evaporation in Florida.  Florida
    Bureau of Geology Map Series 32.

Wedderburn, L.A., M.S. Knapp, D.P. Waltz and W.S. Burns.
    1982.  Hydrogeologic Reconnaissance of Lee County,
    Florida, Part 1-Text.  Technical Publication  82-1.   South
    Florida Water Management  District.

Wilson, W.E.  1977.  Ground-Water Resources of DeSoto and
    Hardee Counties, Florida.  Florida Bureau of  Geology
    Report of Investigation 83.

Wilson, W.E. and J.M. Gerhert.   1982.  Simulated  Effects of
    Ground-Water Development  on  the  Potentiometric Surface of
    the Floridan Aquifer, West-Central Florida.   U.S.
    Geological Survey Professional  Paper 1217.

Wolansky, R.M.   1978.  Feasibility  of Water Supply Development
    from the Unconfined Aquifer  in  Charolotte County, Florida.
    U.S. Geological  Survey Water-Resources Investigation
    78-26.

Wolansky, R.W.,  L.R. Mills and W.M.  Woodman.   1978.  Water
    Table in  Surficial Aquifer,  West Central  Florida.   U.S.
    Geological Survey  Open File  Report No.  78-1045.

                              274

-------
                          APPENDIX A
                      PRZM MODIFICATIONS
One  of  the tasks required by the Florida Teraik study  was  to
make  modifications to PRZM to cover special situations arising
from the Florida hydrologic situation and the use of aldicarb.

PRZM  currently  is a one-dimensional hydrologic and  transport
model  for pesticides in the unsaturated zone.  Aldicarb  forms
toxic  daughter products which may have different environmental
and  toxicological  properties  from  the  parent.   Therefore,
algorithms  to  handle this situation were added.   Because  of
the  variety of irrigation management practices used on Florida
citrus  and  their  potential  impact on  transport  a  set  of
algorithms  were  implemented  to handle  these  circumstances.
PRZM  currently  allows  for restricted vertical  drainage  but
does  not  make allowance for lateral flow under  scenarios  of
temporary  saturated  conditions caused by restricted  vertical
drainage.   In Florida flatwoods soils, the practice of bedding
citrus  necessitates  such an addition.  Such an algorithm  was
implemented.

In  depth  discussions  of  the algorithms  are  given  in  the
following sections.
                              275

-------
A.I  ADDITION OF PARENT/DAUGHTER CONTAMINANT RELATIONSHIPS
The  fate of pesticides in soils is a complex issue.  There are
many  processes (i.e., volatilization, degradation, etc.) which
must  be considered in order to adequately address this  issue.
One  of  these processes, which has been largely  neglected  in
pesticide  leaching  models, is that of the  transformation  of
the  parent  compound to various toxic daughter products.   The
tendency  has  been to lump all the toxic family into a  "total
toxic  residue"  and  model  the fate of this  composite  as  a
single  chemical.   This  assumption  may  not  be  acceptable,
especially  if the daughters have very different decay rates or
adsorption  partition coefficients from the parent or from each
other.

Aldicarb  is  oxidized to a sulfoxide and a sulfone as part  of
the  transformation/decay process.  Algorithms were put in PRZM
to   simulate  parent/daughter  relationships.   An  analytical
solution  to the decay and transformation equation was  derived
to check the numerical model.
The  system which was modeled is shown in Figure A.I.  The   GJ
                                              *                ±
are   dissolved   concentrations  and  the  C ^  are   adsorbed
concentrations.   The K. are adsorption partition coefficients,
the  k .  are  decay and transformation rates in  the  dissolved
       D             *
places  and  the  k^  are adsorbed  phase  decay  coefficients.
Notice  that  only the dissolved forms nay be transformed  from
one   toxic   form  to  another.   A  system  of  first   order
differential  equations  describing this system can be  written
as:
                              276

-------
                      C2*      C3*         ADSORBED PHASE
                            4
                       C2   _3_^ C3          DISSOLVED PHASE
              T   k,    T
1    '   k3
Figure  A.I  Schematic of a system of parent
             and daughter pesticides.
                        277

-------
                                                         (A.2)
                       dt
                                                           .3)
                             *
                         d C1P
                         -at—
                        d C *P       *  *
                         dt
                                                         (A.6)
Making  use  of  C^  Ki = ci the s^x  equations  above  can  be

reduced to three equations in three unknowns,  namely:
                         d C.                             (A.7)
                      d C0

                                                         (A.8)
                             278

-------
in which
                     d C3  _     -                         (A.9)
                      dt   ~ a4 C2 * "5 C3
                                                         (A-
                 al
                                                         (A.11)
                               k 0                        (A.13)
                        a4 =    4
                                                         (A.14)
                              Uf^S ~ ~3 "3P
                     S5
These    ordinary   differential    equations    with    constant
coefficients  can  be   solved  analytically for C^,   €2  and  €3
using  the  initial  conditions,   C^  =  C'-^  when  t  =  0  and
C-= C-.= 0 at t = 0.   The  solutions are:
                              279

-------
C.  =  C.'
                                                         (A. 15)
and
                                       ' e
                 E^ a4
              a3 ' 35)
 - a^    (a^ - ag)
                        mm

                                 (A.17)
In  PRZM,  the equations are solved numerically  as  part  of  the
general  advection-dispersion equation  for  a  solute in a porous
medium  by using a backward difference  implicit  scheme.   A  new
subroutine,  PSTLNK,  was  added  to set up  the   transformation
(source  and  sink)  terms  for the system.    In the  case  of
aldicarb,  we have the  relationship C-^-»-C2^C3,  but the   system
can  be configured for  ^••\^<^  an<^ '"l^^B or
C _  and C ,, simply by  selecting  zero  or  positive  values for the
appropriate transformation  rate  constants.

Figures  A. 2 through A. 4  show  the  results  of a series of  tests
performed  on the numerical model  and checked by  the analytical
model.   In these figures,  the solid  line  represents the "true"

                              280

-------
to
CO
    100
     80
     60
40
     20
                         NUMERICAL

                   	 ANALYTICAL
                                          8        10        12

                                             TIME, IN DAYS
                                                              14
16
                                                                               18
20
           Figure  A. 2   Conversion of GI  to C2 to C3  with no  adsorption  and no decay.

-------
                 100
M
00
N)
              EH
              2
              W
              w
              04
2
O
M
fi


I
2
(0


§
u
                                                                           1    I    I   I

                                                                              —— NUMERICAL

                                                                              •	ANALYTICAL
                                                 8      10       12


                                                    TIME, IN DAYS
                Figure A. 3   Conversion of C,  to C,  to C3  with decay but no adsorption.

-------
              100
00
            H
            Z
            U
            U
            cu
            U
            U
            z
                                                                             NUMERICAL

                                                                        	 ANALYTICAL
               20 -
                                                     10


                                                 TIME, IN DAYS
               Figure  A. 4   Conversion  of C-^  to C- to C-j with decay and  adsorption,

-------
or  analytical  solution, while the dashed line represents  the
approximate  numerical solution.  In Figure 2.26, there was  no
decay  of  the dissolved phase chemicals and no  adsorption  of
any  species.   The  rate of transformation from C^ to  €2  was
0.2/day  and that from €2 to €3, 0.5/day.  After 20 days nearly
all  the  chemical is in form C-j.  The numerical  model  traces
the  decay  and  formation of each constituent  closely,  being
poorer  in  those  regions  where the rate  of  change  of  the
concentrations  are  more  rapid.  Figure A.3  shows  the  same
system  with  a decay rate of 0.01/day in the dissolved  phase.
Figure  A.4 shows the solution for the same system except  that
the  adsorption coefficients have been given values (Ki =  0.5,
K   =  1.0  and  K-. = 5.0) as have  the  adsorbed  phase  decay
 ^              * J
coefficients  (k. = 0.01).  Notice that the transformation rate
is retarded when adsorption is introduced.

Using  the  analytical  model, the assumption of  modeling  the
"total  toxic  residue"  decay  as a  first-order  process  was
tested.   Adsorption coefficients for a Woburn sandy loam (K-, =
0.55,  K   =0.16 and K- = 0.185) and decay and  transformation
rate  constants  (k  = 0.07, k_ = 0.55, k3 = 0.01, k4  =  0.031
and  k,_  =  0.0152) were taken from Bromilow et al.,  1980.   A
      5                                                    33
soil  bulk  density  of  1.45, a water content of  0.27  cm  /cm
and  an initial aldicarb parent mass of  100 mg were also  used.
The  model  was  run for 90 days and the results are  shown  in
Figure A.5.

The  results  show that  the decay of the sum of  the  dissolved
aldicarb  concentrations does not follow first-order  kinetics.
The  reason  for this is the conversion  of aldicarb  parent  to
aldicarb   sulfoxide.    Because  the  sulfoxide  has  a   lower
partition  coefficient,  the dissolved concentration  increases
until  most of this conversion is complete.  Once this happens,
however,   the   sum   of  the  sulfoxide   and   the   sulfone
concentrations  does  follow  a first-order decay  curve.   The
                              284

-------
            120
oo
Cn
                                                                      0 PARENT ALDICARB
                                                                      A SULFOXIDE
                                                                      V SULFONE
                                                                      D TOTAL
                                                40       50

                                                TIME, IN  DAYS
          Figure  A-5  Conversion of  aldicarb to  aldicarb  sulfoxide to  aldicarb  sulfone,

-------
implication  is  that as long as the parent exists, the  parent
and  at  least  the  sum of the  daughters  should  be  modeled
separately.   Thus,  in the unsaturated  zone,  parent/daughter
relationships  should  be  used.   However, if  little  of  the
parent   reaches  the  saturated  zone,  the  sum  of  aldicarb
sulfoxide  and  sulfone  can  be modeled  as  a  single  solute
following first-order decay.
A.2  ADDITION OF IRRIGATION ALGORITHMS
A  new subroutine, IRIGAT, was added to the PRZM code in  order
to simulate the application of irrigation water.

There  are three things which are crucial to the application of
water;  the  time that the event is triggered, the total  water
applied,  and the rate at which it is applied which  determines
the length of the irrigation event.

The   subroutine  was  designed  to  trigger  irrigation  water
application  when the average water content in the root zone is
at  a  user specified percentage of the available water in  the
root  zone (the water content at field capacity (9 £ J minus the
water  content  at  wilting  point (9   )).   The  total  water
applied   at that time, is calculated by:
                           fc. -. 0 ±)A«.                 V-18)
This  is the water required to bring the root zone profile back
to  field  capacity.   Different types  of  irrigation  methods
(e.g.,   overhead  sprinklers  and  trickle)  apply  water   at
different  rates.   The  algorithms allow the user  to  specify

                              286

-------
that  rate.  The length of time over which the water is applied
is  then calculated by dividing the total water applied by  the
application  rate.   This is rounded to the nearest  whole  day
because  PRZM  operates on a daily timestep.   The  application
rate  is  adjusted  accordingly so that the  same  total  water
quantity is applied.
A.3  ADDITION OF LATERAL DRAINAGE ALGORITHMS
In  a soil profile when drainage is restricted there is  always
the  possibility  of  lateral flow occurring if  perched  water
tables  (i.e.,  saturated  soil conditions)  occur.   In  PRZM,
drainage  is restricted by using a time constant which modifies
percolation  from  the  soil layer.  The  equation  is  (Carsel
et al., 1984) :
                                    -
                     __=-   L        fc              (A. 19)
In  this  expression,  9  is the  dimensionless  hydraulic  head
acting  on  the  soil layer (cm water/cm soil) and  8 fc is  the
field capacity water content.  It has the solution:
                                          *             
-------
             e - efj
              .dt rc   = -  a  + i    ( e -  0)          (A.21)
In  this equation, K? is the time constant for lateral drainage
and  is only positive if 6 > 9  where 9  is the saturation water
                             s        s
content  in the layer.  Otherwise, K~ = 0-  The equation has  a
solution similar to (2.21), specifically:
Equation  A.22  is used if 9  >9  and if this condition is  not
                            o   s
met  over the entire time step, the algorithm computes the time
't1   when  the change occurs and switches to (A.20)  for  the
remainder  of the time step.  This is, obviously, a very simple
analytical  model  with empirical coefficients  representing  a
rather  complex  physical  process.  The concept is that  of  a
tank  with a hole in it and another hole in the side very  near
the  bottom.  The quantity of water flowing out depends only on
the   depth   of  water  in  the  tank  ( 9 - 9 ^ )   and   the
characteristics of the holes (K^ and K2)•

Equation  (A.22)  gives  the value of   at any  time  't1.   To
determine  the  quantities  of  water  draining  laterally  and
percolating,  it  is necessary to derive expressions for  these
quantities.  The instantaneous rate of water percolating is:
                                                        (A.23)
                             288

-------
where   Az  is   the  thickness  of  the  soil  layer.   One  can
substitute   the solution for 8 (A.22) into (A.23) and integrate
over   time  to  obtain the total percolation (Q-^) over time  't',
thus:
                                       - (K, +  K,)  t
Q, =  • IAS K,  (e *. +  (e n  - e ff,> e     A     2.   > at
          fc  T l * o   * fc'
o
 This  integral has the solution:
                           _   -'0^+  ^t
                        IT
      .    **Kl(0   -0  fc)     _
   gl ~
 Similarly,  for the lateral flow:
                   ^ o — ^ fe
   Oj        /T-r  i  »» \         I A  ^
 These  equations are subject to the initial conditions Q-^ =  Q2
 =  0, v/hen t = 0.  The time constant for lateral drainage, Kj  /
 can be estimated from available field data.

 Although   this  is  a  simplistic  representation  of   lateral
 drainage  it can be used to bracket the two-dimensional  problem
 of  lateral drainage and percolation with adsorption  and decay,
 since  one  can either allow the lateral flow to exit the  soil
                               289

-------
column  at  zero  pesticide concentration or at  the  dissolved
concentrations in the soil column.
                              290

-------
                   APPENDIX B
Plots of soil order physical properties from soil
            characterization analysis
                      291

-------
VO
KJ
                                                                         y-s
           40. O
           30.0-
         LU
         S 25. 0+
         LU
           20. 0--
            5. 0--
0.0
                                 1	1	1	1
                                              H	1	1	1	1	1	1	1-
                                                                          H	1	1-
0
                             50
100           150

   DEPTH  (CM)
                                                          200
250
        Figure B.I  Mean and standard deviation of field capacity water content  vs.
                   depth for Astatula and Candler entisols.

-------
to
vo
                                                y+s
                                                             y-s
          40.0-



          35. 0;:



          30.0-
       LU
       S 25. 0+
       Dd
       LU
       Q.
20.0-



15.0-
       Q 10.0"
       _j
       LLl
       I—I
       U_

           5. 0-f"
           0.0
          H	1	1
                     H	1	1	1	1	1	1	1	1   I	1	1	1	1  >   I
              0
50
                               100           150

                                  DEPTH  (CM)
200
250
        Figure B.2  Mean and standard deviation of field  capacity water content vs. depth
                   for St. Lucie and Paola entisols.

-------
to
vo
                                                y+S
                                                                y-s
   40.0-




   35-°:



   30.0-
        LU
S 25. 0+

o:
UJ


g 20. 0--





~ 15.0-





Q 10.0"
        Q.
        LU
           5. 0--
           0.0
                                                                    /  v
                      N.
               0
          -t	1	1	1	1	1	1  I	1	1	1  I	1
                     I  I	1	1   I	1	1
                    50
100
150
200
250
                                            DEPTH (CM)
        Figure B.3  Mean and standard deviation of  field  capacity water content vs.

                    depth for Apopka and Arredondo  ultisols.

-------
to
vo
Ul
                                                  y+S
                              y-s
           40. 0T
                             50
100
150
200
250
                                              DEPTH  (CM)
         Figure  B.4  Mean  and  standard deviation of  field capacity water content vs.
                    depth for Iiranokalee, Oldsmar, Wabasso and Myakka spodosols.

-------
                                                 y+S
K)
vo
en
                              y-s
           40. 0T
                             50
100           150

   DEPTH  (CM)
200
250
        Figure B.5  Mean and standard deviation of field capacity water content  vs.  depth
                   for Felda, Riviera and Pineda alfisols.

-------
                                                y+s
           4. Or
           3.0-
                                                   y-s
vo
        O
        DQ
        o:
        u
           2. 0--
           1.0-
          0.0-1
              0
        Figure  B.6
       50
100           150

   DEPTH  (CM)
200
250
Mean and standard  deviation of organic carbon vs.  depth  for Astatula
and Candler entisols.

-------
                                                y+s
                               y-s
to
vo
00
          4. 0T
          3.0-
        a
        CO
        a:
          2. 0--
             0
100
150
200
250
                                            DEPTH  (CM)

       Figure B.7  Mean and standard deviation of organic carbon  vs. depth  for St. Lucie
                   and Paola entisols.

-------
                                                y+S
y-s
to
vo
vo
          4.0T
             0
                                            DEPTH  (CM)
            250
        Figure B.8  Mean and standard deviation of  organic  carbon vs. depth for Apopka

                    and Arredondo ultisols.

-------
o
o
                                                y+s
                              y-s
          4. 0T
          0
                           50
100
150
200
250
                                            DEPTH  (CM)

         Figure B.9  Mean and standard deviation of organic carbon vs.  depth for
                     Immakolee,  Oldsmar, Wabasso, and Myakka spodosols.

-------
                                         y+S
   4. OT
       .. \
         \
           \

   3.0+    I
 CD
 m
 ct

 u 2.0+

 o
CD
a:
o
   1.0--
             £-H	nJ=q	,	p=H	,	1	,	,—H
50
                                  100           150


                                     DEPTH  (CM)
   y-s
200
250
Figure B.10  Mean and standard  deviation of organic carbon vs. depth for Felda,

             Riviera, and  Pineda  alfisols.

-------
U)
o
9. Oj


8.0--


7.0-


6.0


5.0-


4.0-


3.0-
          1.0-
          0.0
                                              y+S
                                                           y-s
      H	1	1	1
                                                    H	1
             0
                50
100          150

   DEPTH  (CM)
200
250
         Figure B.ll  Mean and   standard deviation of pH vs. depth  for Astatula and
                     Candler entisols.

-------
U)
o
U)
                                               y+s
                                                         y-s
          9. O


          8.0


          7.0


          6. 0


          5.0
          3. Oy


          2.0-


          1.0-
          0.0
\-\—.—I—i—I—h-
 0             50
      H	1	H
                                         100           150

                                           DEPTH (CM)
200
250
          Figure B.12  Mean and standard deviation of pH vs.  depth  for St. Lucie and
                       Paola entisols.

-------
U)
o
                                             y+S
                                                            y-s
      Q_
9. 0T


8.0


7.0


6.0


5.0


4.0'


3.0-


2.0-


1.0-
        0.0
      •4	1	1
           0
                                               1	1	H
                50
100           150

   DEPTH  (CM)
H - »
            ' - 1
200
             250
       Figure  B.13  Mean and standard  deviation of pH vs.  depth for Apopka and
                   Arredondo ultisols.

-------
                                        y+s
                              y-s
   9. 0T
                   50
100
150
200
250
                                    DEPTH  (CM)
Figure B.14  Mean and  standard deviation of pH vs. depth for Immokalee,  Oldsmar,
             Wabosso and Myakka spodosols.

-------
U)
o
                                               y+S
                                                           y-s
         0.
9. O


8.0


7.0


6.0-


5.0-


4.0;-


3.0-


2.0-


1.0-
           0.0
           H	1-
                       1	1	1
                                               1	1	1	H-
                                                            H	1	1   I
              0
                50
100           150

   DEPTH  (CM)
200
250
        Figure B.15  Mean and standard deviation of pH vs. depth for Felda, Riviera and
                     Pineda  alfisols.

-------
                         APPENDIX  C

           DERIVATION  OF  FIRST-ORDER HYDROLYSIS  RATE
                 EQUATION FOR SATURATED ZONE
Hydrolysis  of  aldicarb  is  strongly dependent  upon  pH  and
temperature.   This  reaction  is a typically first  order  and
that follows the equation:
                                                         (c.i)
where    [A]  = concentration of aldicarb (moles/liter)
          k  = 1st order rate constant (/time)
          t  = time

The  second  order  rate  constant (kj. ) can  be  determined  by
plotting  first order rate constant  (kt) versus pH from several
different reactions:
                       k  = —^-                         (C.2)
                        r   [OH-]
The slope of the line is the second order rate constant  (kr).

In  the  late 1800's Arrhenius observed that in  most  chemical
reactions   the  rate  constant  increases  exponentially   with
temperature.    He   described  this  relationship   with    the
following empirical equation:

                              307

-------
                 kr =  A exp (-EA/RT)                      (c.3)
where     k  = rate constant
           r

          A  = pre-expotential factor (mole/L-t)

          E  = activation energy (energy/mole)
                                                   o
          R  = universal gas constant (energy/mole  K)
                            o
          T  = temperature,  K




The  logarithmic  form of this equation is the equation  for  a

straight line:
                   kr = 2.303RT+
where log k   = dependent variable
                          -E.  1
                                       *                 (C.4)
          1/T = independent variable


          E   = slope
           A

      log(A)  = intercept
Although  this  relationship is empirically derived, a  similar


relationship  can  be  rigorously derived  from  thermodynamics

involving  the  equilibrium  constant  (keg) and the  change  in


free energy (AH°):
                                                          (c>5)
                        dT      RT2
Substituting



              keq =  r   and 4H° = El  -  E2

                     K2

                              308

-------
gives

                d In(k1/k2)
                     dT     "    RT2                     (C'6)
or in general
                     d In (JO    E_
                     	L_ = JL_                      (C.7)
                        dT      RT2
Integration leads to
                       kr= Ae-                           (c.8)
Arrhenius   defined  activation  energy  (E^ )  as  the  energy
threshold  that  molecules must overcome if they are  to  react
with  one another.  If the energy of the molecules is less than
the activation energy no reaction will occur.

The  Arrhenius  equation  is used in this study  to  develop  a
general  equation that relates pH and temperature to the  first
order  reaction rate constant.  Two studies (Lemley and  Zhong,
1983   and   Lemley  and  Zhong,  1984)  have  calculated   the
activation  energy and shown that the log-linear  relationships
necessary for its validity hold.

The  equation  is  rearranged  so  that  by  knowing  EA ,  the
activation  energy  and  A, the pre-exponential  factor  for  a

                              309  :

-------
given  compound,  the  first  order rate  constant  k.  can  be

calculated for any given temperature and pH:
                  kr = 2.303RT
                    kt
Substituting  k  - ——  gives
                  [OH J
                kt \     -E   1
         log (	£-) =	+ log A
               [OH ]/   2.303R T
or
                      -E   1
           log k  =	+ log A + log [OH ]
                r   2.303R T
Substituting  pH - 14 for log .[OH ]   gives the final useful

form
                       -E   1
            log k.  =	+ log A + pH - 14
                     2.303R T
Uncertainty in Reaction Rates


A  great degree of care must be taken  in how the reaction  rates

k    are determined and used.  Large uncertainty is   introduced

from several different sources:


                             310

-------
    1)   temperature sensitivity
    2)   large extrapolations
    3)   experimental errors, and
    4)   environmental fluctuations.

Small    temperature   fluctuations   during   the   hydrolysis
experiments  can lead to large error propagation in k  and E, .
A  variation  in  temperature  as small as .2  C  leads  to  2%
uncertainty  in  k   and  5% in E .  With a + 1° C  change  the
                  t              A
uncertainty  in  k  is 10% and in E ,  100%.  Larger  variations
                  t                A
in  temperature of 10  and 25  will translate to uncertainty in
k  on the order of a factor of 2.5 and 10 respectively.

Extrapolation  over large temperature or pH intervals can  also
lead  to error.  With an uncertainty in E  of 5%,  extrapolation
                                         f\
over  a  25°C temperature interval can result in 30%  error  in
the estimation of kf
Experimental  error  is  another source of  uncertainty.   Most
rates  are based on only three to six data points.  With such a
small  sample  size  the influence of one outlier or  bad  data
point can skew the entire study.

The  rate constants are usually used to predict degradation  in
the  natural  environment where there are temporal and  spacial
fluctuations of both temperature and pH.

Due  to all these sources of error, it is best to only consider
the estimate of k  as being within an order of magnitude.
                              311

-------
                          APPENDIX D
Aldicarb TTR data from Oviedo and Davenport sites for 1984,
                              312

-------
                                        Davenport 5/9/84


QUAD       CORE #            DEPTH(FT)       TTR (ppb)



 NW           2                0-16          -ND


 NW           8                0-16          -ND
 NE          13                0-1            ND
                               1-2            ND
                               2-4            ND
                               4-6            ND
                               6-8            60
                               8-10          144
                              10-12          72
                              12-14          ND
                              14-16          ND
 NE
0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
28
ND
ND
ND
ND
44
100
216
40
 NE           12                0-1            32
                               1-2            ND
                               2-4            ND
                               4-6            ND
                               6-8            52
                               8-10           88
                              10-12           56
                              12-14           84
                              14-16           56
  SW          15                0-16          -ND
                         313

-------
                              Davenport 5/9/84






QUAD       CORE #            DEPTH(FT)      TTR (ppb)
SW 3 0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
SE 5 0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
SE 13 0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
76
ND
ND
ND
ND
ND
16
52
96
84
ND
ND
32
76
104
80
72
44
64
ND
36
44
64
92
100
212
48
                        314

-------
                               Davenport 6/13/84


QUAD       CORE #            DEPTH(FT)       TTR  (ppb)



 NW           3                0-16           ND


 NW           8                0-16           ND
 SW          11                0-1            ND
                               1-2            ND
                               2-4            44
                               4-6            104
                               6-8           296
                               8-10          200
                              10-12          328
                              12-14          272
                              14-16           88
 SW
0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
ND
ND
24
48
32
32
28
12
16
 NE           6                0-1             16
                               1-2             ND
                               2-4             22
                               4-6             40
                               6-8             80
                               8-10          120
                              10-12           84
                              12-14           92
                              14-16           96
                         315

-------
                              Davenport 6/13/84






QUAD       CORE #           DEPTH(FT)       TTR (ppb)
NE 12








NE 13








SE 4








SE 13








0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
0-1
1-2
2-4
4-6
6-8
8-10
10-12
12-14
14-16
36
28
28
60
116
152
168
84
ND
16
ND
28
40
72
92
60
28
20
ND
ND
20
40
52
72
112
164
152
16
12
16
56
60
60
84
88
52
                        316

-------
                              Oviedo 5/22/84






ROW         COL.             DEPTH(FT)       TTR (ppb)
11-12 38




11-12 34




11-12 13




13-14 17




13-14 52




25-26 53




21-22 50




0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
1460
284
ND
ND
ND
ND
ND
ND
ND
ND
80
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
580
76
128
368
ND
ND
ND
ND
ND
ND
                       317

-------
                              Oviedo 5/22/84






ROW         COL.             DEPTH(FT)       TTR(ppb)
24-25 42




22-23 31




27-28 28




28-29 29




28-29 15




27-28 12




18-19 4




0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
32
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
544
116
ND
ND
ND
ND
ND
ND
ND
ND
                      318

-------
                               Oviedo  5/22/84


 ROW         COL.             DEPTH(FT)      TTR (ppb)
40-41         48                0-1             80
                               1-2             ND
                               2-3             ND
                               3-4             ND
                               4-5             ND

32-33         29                0-1             ND
                               1-2             ND
                               2-3             ND
                               3-4             ND
                               4-5             ND
                         319

-------
                              Oviedo 6/20/84







ROW         COL.             DEPTH(FT)       TTR (ppb)
12-13 42




13-14 6




17-18 15




17-18 15




18-19 26




34-35 - 48




30-31 33




0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1 (iced)
1-2
2-3
3-4
4-5
0-1 (no ice)
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
0-1
1-2
2-3
3-4
4-5
44
ND
28
ND
ND
ND
ND
ND
ND
ND
60
12
288
84
ND
80
12
40
76
ND
ND
ND
ND
32
ND
ND
ND
ND
ND
48
ND
ND
ND
ND
ND
                        320

-------
                               Oviedo 6/20/84


 ROW         COL.             DEPTH(FT)       TTR (ppb)
26-27 20(#1)




0-1
1-2
2-3
3-4
4-5
ND
ND
24
ND
ND
26-27    .     20(#2)            0-1             ND
                               1-2             ND
                               2-3             ND
                               3-4             ND
                               4-5             ND
                         321

-------
                          APPENDIX E
   CUMULATIVE FREQUENCY DISTRIBUTIONS OF RELATIVE WELL-WATER
                        CONCENTRATIONS
The  following  summary  table  (Table  E.I)  and  graphs  show
cumulative  frequency distributions of relative  concentrations
in well water for the saturated zone scenarios.

The  following notation convention is used for identifying each
scenario.


-------
                     Well Configurations
                        W (distance (ft), depth (ft), rates
                          (gpm))

FWD(0)W(300,50,2000),  for example, would identify the scenario
that  uses  the unconfined Floridan Aquifer with the worst  set
of  hydrogeologic properties, no decay rate, a well distance of
300  ft  from the aldicarb source, a well depth of 50  ft,  and
pumping rate of 2000 gallons per minute  (gpm).
                             323

-------
TABLE E.I   RELATIVE CONCENTRATION VALUES  FOR
              INDICATED  SATURATED SCENARIOS  AT THREE
              EXCEEDANCE PROBABILITIES
Scenario


Floridan Aquifer - worst Case

  91 meters to well
    Deep well, no decay
    Decay
    Shallow well, no decay
    Decay

  300 meters to well
    Deep well, no decay
    Decay
    Shallow well, no decay
    Decay
    Exceedence Probability
 .90         .50         .10
6.4E-4
8.1E-5
4.0E-3
1.3E-4
1.1E-3
3.3E-5
4.0E-3
3.3E-5
4.4E-3
3.7E-4
l.OE-2
5.8E-4
5.0E-3
1.4E-4
8.5E-3
1.8E-4
8.6E-3
1.7E-3
1.5E-2
2.6E-3
8.4E-3
6.2E-4
1.2E-2
6.4E-4
Floridan Aquifer  -  average case

  91 meters to well
    Deep well, no decay              5.7E-4      3.3E-3      7.IE-3
    Decay                           2.5E-6      1.8E-5      4.5E-3
    Shallow well, no decay           6.3E-4      3.6E-3      7.5E-3
    Decay                           2.5E-6      1.8E-5      4.4E-5

  300 meters to well
    Deep well, no decay              1.6E-4      7.3E-4      3.0E-3
    Deep well, decay                 9.2E-9      3.1E-8      4.6E-8
    Shallow well, no decay           9.6E-5      4.1E-4      1.9E-3
    Decay                           2.6E-9      8.7E-9      1.3E-8
Surficial Aquifer  - worst case
  91 meters to well
    High pumping rate, no decay
    Low pumping rate, no decay
    Both, decay

  300 meters to well
    High pumping rate, no decay
    Low pumping rate, no decay
    Both, decay
1.1E-2
2.8E-2
2.1E-4
5.4E-3
6.7E-3
4.0E-6
5.0E-2
1.3E-1
1.1E-3
2.6E-2
4.7E-2
1.8E-5
7.3E-2
1.7E-1
3.3E-3
5.9E-2
l.OE-1
3.3E-5
Two-aquifer System, worst case
  91 meters to well
    High pumping,  no  decay           1.5E-4      6.9E-4      1.7E-3
    Low pumping, no decay            3.4E-4      1.1E-3      2.1E-3
    Both, decay                     1.3E-5      6.5E-5      2.7E-4
                                324

-------
a o
i— § •

se
a:
O a

"s
LJ
£-1 •


5
tiJ
or c
   •

  a.
                    FWD(0)M(300,10-40,200)
                       FWD(0)H(300,10-40,40)


                  \J*
    0.00   0.25
0.50     0.75     1.00     1.25

  RELflTIVE  CONCENTRRTION
1.50
1.75

*io-2
                            325

-------
o
LJ O

|S

LiJ
   O

i—« {5
o o
a:
E-< a
Sg-
O o
"o'J
   in
      \
 .
CC  O
{-•
"ZL a

o 2'
a: "*
LJ
a. °
   a.
       \
        \
         \

         \
          v

          \
           \
            \

             \
\
 \

 \

  i
CiJ
   O                           ~Ns.
    .                              x

                                              FWD(29.7)W(300,10-40,40)
                                          «».
                                             i
                        FMD(29.7)W(300,10-40,200)
     0.0     5.0      10.0      15.0     20.0     25.0     30.0     35.0
                       RELflTIVE CONCENTRATION
                                 326

-------
XN ^ PWD ( 0 ) W( 300 , 50- 350 , 1000 )
              FWD(0)W( 300, 50-350, 500)
0.0000.001  0.002
  	    I      I      	    	    	
0.003  0.004 0.005  0.006  0.007  0.008

 RELflTIVE CONCENTRflTION
                                 0.009  0.010
       327

-------
LJ -
a
LJ o
LJ
   o
»—i 8
O o
So
O o
O
   o-
   in
•—« o
5*
cr: c
^ o J
E-
"Z. o
^2J
  0
Q- °.
  o.

       \

        \

        \
         \
         \
         \

           \
            \

            \
             \

              \
               \
               \
                \
                    N
                     N
                                           FWD(29.7)H(300,50-350,500)
                         FWD(29.7)W(300,50-350,1000)
     0.0     2.5     5.0     7.5     10.0    12.5    15.0     17.5    20.0
                      RELATIVE CONCENTRRTION
                                328

-------
  a

a §-

a
CJ O

gg-

u
  o
cn_-_
O a
E-« o


§8"


O o
  *-
(C O




E-i
     \.
                               FWD(0)W(1000,10-40,200)


                                FWD(0)W(1000,10-40,40)
     0.000  0.002
                   0.004    0.006    0.008     0.010

                      RELflTIVE  CONCENTRflTION
0.012
0.014
                               329

-------
s§
a
LJ o
8s
O O
a:
£-• 0

§38
o
2
O o
   tn
LJ
>
i— i a
LJ
o
z
a: a

    -
0- °.
  o.
     \

      \

       \

       X
        \

         \
         \

          \

          \
           \

           \

            \

             \
             \
             \


              \
               \
               v
   a

                                   "X.
                                            FHD(29.7)W(1000,10-40,40)
                                     K
                                     I      ^
                FWD(29.Z)W(1000,10-40,200)        \
•z. o

o2'                                                -N
a:                                                    •
                                                      \
                                                        \
     0.0    1.0     2.0     3.0     4.0     5.0     6.0     7.0     8.0
                      RELRTIVE  CONCENTRRTION
                                 330

-------
       1000,50-350,1000)
           FWD(0)W(1000,50-350,500)
2.0    3.0    4.0    5.0    6.0    7.0
     RELflTIVE  CONCENTRflTION
8.0    9.0
      HlO"
               331

-------
   o
LJ
cn Q.
•—« a
Z
O C3

o:
£-« o
s§-

O o
I—I o
SsH
o:
cn o
£-1
•2L O
o:
\
\
 \
  X
  \
   \
   \
         \
          \
          \
          \
           \
            \
            \
             \
              \
              \
                           FHD(29.7)W(1000,50-350,500)
          FHD(29.7)H(1000,50-350,1000)
                                            \
     0.0     1.0     2.0     3.0     4.0     5.0     6.0     7.0     8.0
                      RELRTIVE  CONCENTRRTION               xlO~*
                                 332

-------
3s
  o
o
CJ

08
oa

QJ
  O
O o



o:
f-" o

SB"'


O a



>
>-i o
_J
LJ
ct: a
a: c
T*  •



£-•
Z a
o:
CJ
     \
      v
          \
           \\
Vx  FAD(0)W(300,10-40,200)


 Vv/  FAD(0)W(300,10-40,40)
             \\
                \
                  \
                   \
                     \.
                              \\
                                           \\
    0.000 0.001  0.002  0.003  0.004  0.005  0.006  0.007  0.008  0.009

                  RELRTIVE CONCENTRATION
                          333

-------
CxJ O


8 B-
X
LJ
   a
O o
    •
a:
E-i
   8-
O o

"oJ
 . if>
       \
       \
        \
        \
         \
          »

               «

CiJ


02 °    FAD(29.7)W(300,10-40,200)


O "


en o
Z o

8 a-
o:
                                                x FAD(29.7)H(300,10-40,40)
                                                \
                                                 \



                                                 \
                                                   \

                                                     N
                                                     N
                                                      N
     0.0  5.0   10.0  15.0  20.0   25.0  30.0  35.0  40.0   45.0  50.0  55.0
                        RCLRTIVE CONCENTRflTION
                                   334

-------
 O
 CJ O
O o
   s

O o
»-« o
CO
or
   B-
cc a

OR'
0.0
  O.
\ \^FAD(0)W(300,50-350,500)
   \  V
                             \
                                \
               ,^FAD(0%)W(300,50-350,1000)
                                                   \
     0.0   1.0    2.0    3.0     4.0    5.0    6.0    7.0     8.0    9.0
                       RELRTIVE CONCENTRATION
                                              MlO
                                 335

-------
o
CO o
CO QV

08


CO
   O


i—i 8"



O a
CC

E-. o
  8'            .     .

                              FAD(29.7)W(300,50-350,500)

Oo
•— i o



§*'
u
ce ca
CC CJ

OR'-
Z o

Ss-
a:
LJ _
a. °.
  o.
     \

      \

       \
       \
         \

         \
          \

           \

           \

                     \
                      \
          FAD(29.7)W(300,50-350,1000)
     0.0  5.0  10.0  15.0  20.0   25.0  30.0  35.0  iO.O  45.0   50.0  55.0


                      RELRTIVE  CDNCENTRRTION               *10~"
                                 336

-------
Q 5
CJ ""
a
cj o
o a
          t
a:
18'

O o

-------
S §~Ts*
O
  O
CO              X
   O


Kg-              \   \

                      \







                       \\

                          \       \      FAD(29.7)H(1000,10-40,40)

o -


a o                         \

   d-
   tn                           \
                               \
                                \
                                \

                                '\
                                  \
                                  \
                                  \
                                   \
                                   \




8 "1


33°. J

oRl
                                        \
                                        \
                                        \
                                         \
                                         \
                                         \
                                          \
                                          \
                                          \
                                           \



     0.0      2.5      5.0      7.5      10.0      12.5     15.0      17.5
                             /
        FAD(29.7)H(1000,10-40,200)
                      RELRTIVE CONCENTRRTION
                                338

-------
   o

gS



|i
         \

CD o
   8
O o
ct: a
a:
CJ
      i
        \


             \
              \
                N\
                     FAD(0)W(1000,50-350,500)




                                      50-350,1000)
                                             \
                                                       '^^
     0.0   5.0   10.0   15.0   20.0   25.0   30.0   35.0   40.0   45.0

                       RELRTIVE CONCENTRflTION               *10"*
                                 339

-------
0§-N.
to -
a
LiJ O

I"-


"
00
01
FH 0
a o
o 0'.
  in
>
•—« o
a: a
f-
z o
Ss-
a:
                  .


                  \\
                       \
                       s
                        \
                         N
                         . S
                           FAD(29.7)M(1000,50-350,500)
                          J
                       r
      FAD(29.7)H(1000,50-350,1000)
N
N
 S
 \
                             V

                              N
                               \
                               \
                               \
                                N
                                \
                                \
                                 \
                                 \
                                  \
                                  \
                                   \
                                   \
                                   \
                                    X
                                    \
                                    \
                                     N
                                     \
                                     \
                                      \
    0.0  5.0  10.0  15.0  20.0  25.0  30.0  35.0  -40.0  45.0  50.0  55.0
                 RELRTIVE CONCENTRRTION
                        340

-------
LJ
  a
"J r-».
08


LJ
  O

CO -I.
O 0
a:
ct:
           \
            \
              \
   *
|B


O o

°9
WS
*" 9"


a: o'
   •


o

cc o





Z a

B°H

      \


                    X
         \


                      \^SHD(0)W(300, 10-35,40)

\           " --\
   \                   \
     V-*SWD(0)W(300,10-35,200) \
     \
                                         \

                                         \

                                          \

                                          \



                                          \

                                           \

                                           \

                                           \


                                            N
                      \



                      \
                                               \
                                                \

                                                 \
                                                 N
                                                  \
                       V
                                                    \
    0.000 0.025   0.050  0.075   0.100   0.125   0.150   0.175   0.200
                   RELRTIVE CONCENTRflTION
                           341

-------
tlJ
Q
U o
tJ rj.
08
O o
HH  '

£s'
cc
\
O
z
O o

"o-
   in
CC

QJ
a:
                                          SHD(29.7)W(300,10-35,200)
                                                 SWD(29.7)H(300,10-35,40)
                                      "•.—.v......
Z o

y°J
ttT
CJ _
0.0
5.0
10.0
15.0
20.0
25.0
30.0
                       RELflTIVE CONCENTRRTION
                                          35.0
                                                                40.0
                                 342

-------
     V
LJ O
wx-  \v

  a
O o
en
E-« o
18-
O o

S8''
«—« o
e*
QJ
a:
az o
       \
       NX
        \\
          \\
           *.  \
           \ \
            \ \
             X
             \
                        SWD(0)H(1000,10-35,40)
              •
                 \
                   \
                      \
  B-
                          \ ^rSWD(0)W(1000,10-35,200)
                           ,
                            \
                             \
                              \.

                               \
    0.00    0.02     0.04     0.06      0.06      0.10      0.12
                    RELRTIVE CONCENTRflTION
                             343

-------
cj -
o
CiJ a

  *

  -j    ""^
CO ^J       \  \
            \  \
O o
a:
{-• a
68

2
O o
QJ
z
CC C3
2 o

8s-
      \x



                      *-.\
                       \
                                         SWD(29.7)W(1000,10-35,40)
                                  \
                                   \
                                   \
                                    \
                                    \
                                    \
                                     \
                                     \
             SHD(29.7)W(1000,10-35,200)     \
                                        \
                                         \  \

                                           Vx
                                        	\
    0.0   5.0    10.0    15.0   20.0    25.0   30.0    35.0    40.0
                    RELRTIVE CONCENTRRTION
                             344

-------
a 8
a-
a
CO O
LJ
O O
.X.

CJ

»-* a
a: o

  8-
(E O

5s-
  .   v^


  \  x\
     \     \ £• 2WD(0)W(300,50-400,700)
               \     \
                          %
                   (^«2WD(0)W(300,50-400,1500)
                     •»..
                      X     V
                              \
                                            \      \
                                              \
     0.0   2.5    5.0    7.5    10.0   12.5   15.0   17.5   20.0   22.5

                     RELRTIVE CONCENTRRTION              *10~
                              345

-------
  o


°§\
p ^ *^ *
y   *
a   \
  o \



%+ \
CJ     \
  o

£g-


O a     \

          \



  a

  8

82
u1"

               \\
                \  \
                 *.   N.
d               \  \               2WD(29.7)W(300,50-400,1500)
a: o               '	A	•»	x

                                 ^  I
                                  N  ./
                           ^\    \ *
                           V? '•     x


                           r\   \
                                          \
        2WD(29.7)H(300,50-400,700)



S I'                                   "    V


B                                         \   \
CL °.                                         \   \
  O	     V	


    0.0     5.0     10.0    15.0    20.0     25.0    30.0    35.0


                   RELRTIVE CONCENTRRTION
                           346

-------