&EFA
           United States
           Environmental Protection
           Agency
            Environmental Research
            Laboratory
            Athens GA 30605
EPA-600," 5-80-002
January 1980
           Research and Development
Environmental and
Economic Impact of
Agricultural  Land
Use Conversion

An Evaluation
Methodology

-------
                RESEARCH REPORTING SERIES

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

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

This  report has been  assigned  to the SOCIOECONOMIC ENVIRONMENTAL
STUDIES series. This series includes research on environmental management,
economic analysis,  ecological impacts, comprehensive planning  and fore-
casting, and analysis methodologies. Included are tools for determining varying
impacts of alternative policies; analyses of environmental planning techniques
at the regional, state, and local levels; and  approaches to measuring environ-
mental quality perceptions, as well as analysis of ecological and economic im-
pacts of environmental  protection measures. Such topics as urban form, industrial
mix, growth policies, control, and organizational structure are discussed in terms
of optimal environmental performance. These interdisciplinary studies and sys-
tems analyses are presented in forms varying from quantitative relational analyses
to management and  policy-oriented reports.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                             EPA-600/5-80-002
                                             January 1980
     ENVIRONMENTAL AND ECONOMIC IMPACT OF
       AGRICULTURAL LAND USE CONVERSION
           An Evaluation Methodology
                      by
                Fred C. White
              James E. Hairston
               Wesley N. Musser
                H. F. Perkins
                 J. F. Reed
            University of Georgia
            Athens, Georgia 30602
            Grant No. R804510-01
              Project Officers

              Thomas E. Waddell
              George W. Bailey
Technology Development and Applications Branch
      Environmental Research Laboratory
            Athens, Georgia 30605
      ENVIRONMENTAL RESEARCH LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
     U.S. ENVIRONMENTAL PROTECTION AGENCY
             ATHENS, GEORGIA 30605

-------
                                 DISCLAIMER
     This report has been reviewed by the Environmental Research Laboratory,
U.S. Environmental Protection Agency, Athens, Georgia, and approved for publi-
cation.  Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does men-
tion of trade names or commercial products constitute endorsement or recom-
mendation for use.
                                      ii

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

     Agricultural sources contribute significantly to water pollution
problems in many areas of the United States.  These problems are of
particular concern as more farm land is brought  into production to  meet
the demands of expanding domestic and world markets.  This report presents
a methodology for analyzing the environmental and economic impacts  of
incresed crop acreage and applies it to the state of Georgia.  The
methodology was developed as a tool for decision-makers  involved in
formulating public policy concerning levels of agricultural production as
well as evaluating environmental standards for agriculture.

                                  David W. Duttweiler
                                  Director
                                  Environmental Research Laboratory
                                  Athens, Georgia
                                    111

-------
                                  ABSTRACT


     This project, a joint  effort of the Agricultural Economics and Agronomy
Departments at  the University  of Georgia, was designed to develop and apply a
methodology for evaluating  the environmental and economic impacts of placing
marginal, submarginal, and  Soil Bank land in crop production.  Environmental
impacts were measured by quantifying the increased environmental loadings of
sediment, nitrogen and phosphorus and by estimating the potential environ-
mental impact of pesticides.   Economic  consequences considered include changes
in net farm income and aggregate impact on the staters economy.  Although no
attempt was made to place a dollar value on environmental consequences, these
impacts can be  weighed against the economic impacts in a tradeoff fashion as
a measure of social costs and  benefits  resulting from increased production on
new cropland.

     The methodology developed in this  study was applied to  the state of
Georgia for the period 1973-1976,  During this period 74,5 thousand hectares
of marginal, submarginal, and  Soil Bank land in Georgia were placed in crop
production.  This change represents only a small fraction of land in these
categories that could potentially be converted to crop production.  If these
land use changes prove to be profitable over a period of years, thereby vali-
dating the conversion decisions, more of this land would be  expected to be
converted to crop production in the future.  The pollutants  generated annually
from each hectare of this land, on the  average, are estimated to be 11.9 met-
ric tons of sediment, 8,1 kilograms of  nitrogen loss (excluding leaching),
and 10.5 kilograms of phosphorus loss.  Net farm income increased an estimated
$61 per hectare annually.   Aggregate personal income for the state was esti-
mated to increase $2 for each  one dollar increase in net farm income.  Meth-
odology used in this study  could be used in similar studies  in other areas of
the United States.  Much of the required data is already available in other
areas.

     This report was submitted in fulfillment of Grant No. R804510-01 by the
University of Georgia under the sponsorship of the U.S. Environmental Protec-
tion Agency.  This report covers the period July 1, 1976, to August 31, 1978,
and work was completed as of August 31, 1978.
                                      iv

-------
                                  CONTENTS
Foreword	   iii
Abstract	    iv
Figures	   vii
Tables	viii

     1.  Introduction	    1
              Ob j ectives	    2
              Organization of the Study	    2
     2.  Summary and Conclusions	    4
              Methodology	    4
              Empirical Application	    6
     3.  Methodology	    8
              Land Resource Inventory	    8
              Land Use Changes	    13
              Conversion of Soil Bank Land	    15
              Input Usage	    16
              Environmental Impacts	    17
              Economic Impacts	    26
     4.  Land Resources and Use Patterns in Georgia	    32
              Land Resource Areas	    32
              Land Use Patterns	    35
              Soil Bank Land	    37
     5.  Estimation of Changes in Land Use and Input Usage j.n Georgia...    40
              Changes in Crop Acreage	    40
              Shifts in Land Use Patterns	    40
              Land Use Changes by Capability Class	    44
              Changes in Input Usage	    47
     6.  Estimated Environmental Impact of Land Use Changes in Georgia..    50
              Sediment and Runoff Data	    50
              Runoff and Sediment Yield	    54
              Fertilizer Loss	    58
              Pesticide Loss	    58
     7.  Estimated Economic Impact on Georgia	    71
              Net Farm Income	,	    71
              Aggregate Economic Effects	    71

Bibliography	,	    75
Appendices

     A.  LANDSAT Remote Sensing to Determine Land Use Changes	    84
     B.  Interindustry Model of Georgia's State Economy	    87

-------
                                                                   Page

        Related Literature	   87
        Conceptual Framework	   39
        Model Specification	   gg
        Data	   91
        Results	   91
C.  Land Resource Inventory for Georgia	   94
D.  Runoff and Sediment Yield Area	  105
E.  Environmental Impact on All New Cropland in Georgia	  112
                                vi

-------
                                   FIGURES


Number                                                                   Page

  1   Overview of methodology	   9

  2   Schematic showing how land resources were inventoried	  10

  3   Schematic showing identification of land use changes	  14

  4   Schematic showing evaluation of economic and environmental
        impacts associated with increased cropland	  18

  5   Land resource areas of Georgia	  33

  6   Schematic showing calculation of annual potential direct runoff...  51

  7   Schematic showing calculation of sediment loading from surface
        erosion	  52

  8   Average monthly precipitation for Georgia	  69

  9   Median discharge for three representative gauging stations in
        Georgia	  70

 Dl   Slope-effect chart	113
                                     vii

-------
                                   TABLES
Number                                                                    Page

  1   Total Cultivatable, Marginal, Wetlands  and  Submarginal  Lands
        in Georgia by Land Resource Area	   34

  2   Cultivatable, Marginal, Wetlands and Submarginal Lands  in
        Georgia Based on Dominant Capability  Classes Assigned to
        Soil Resource Groups	   36

  3   Conservation Reserve Land in Georgia in 1956-1959 by Land
        Resource Area	   38

  4   Cultivatable, Marginal, Wetlands and Submarginal Land Under
        Conservation Use Only in 1967	   39

  5   Changes  in Land Area of Five Major Row  Crops in Georgia, 1973-
        1976	   41

  6   Estimated Conversion of Woodland, Pasture and Other Land to
        Cropland, Georgia 1973-1976	   42

  7   Status of Pine Plantations Established  in Georgia Under the
        1956-1960 Soil  Bank, Conservation Reserve Program	,....   43

  8   Estimated Land Area Converted to Crop Production by Land
        Resource Area,  Georgia 1973-1976	   46

  9   Representative Input Use and Costs Per  Hectare of Corn  for
        Georgia, 1976	.-.-	   48

 10   Representative Input Use and Cost Per Hectare of Soybeans
        for Georgia, 1976	,	,	,	   49

 11   Estimated Runoff  from New Cropland on Marginal, Submarginal
        and Soil Bank Land in Georgia, 1976		   55

 12   Estimated Increased Sediment Yield from New Cropland on
        Marginal, Submarginal, and Soil Bank  Land in Georgia, 1976	   56

 13   Estimated Sediment Yield from New Cropland on Marginal,
        Submarginal, and Soil Bank Land in Georgia, 1976	   57
                                    viii

-------
Number                                                                   Page

  14   Estimated Nitrogen Loss from New Cropland on Marginal, Sub-
         marginal, and Soil Bank Land in Georgia, 1976 ....... , ..........   59

  15   Estimated Phosphorus Loss from New Cropland on Marginal,
         Submarginal, and Soil Bank Land in Georgia, 1976 ......... , .....   60

  16   Economic Pests of Major Concern for Corn and Soybeans in
         Georgia [[[   61

  17   Selected Characteristics Related to Potential Environmental
         Impact of Pesticides Used on Corn and Soybeans in Georgia,
         1976 ...................................... .....................   62

  18   Application Information, Transport Mode, and Biomagnification
         of Selected Pesticides Capable of Environmental Impact in
         Georgia, 1976 ............ ,...,., ...............................   66

  19   Amount of Pesticides With Potential Environmental Impact
         Applied to New Cropland on Marginal, Submarginal and Soil
         Bank Land by Land Resource Area in Georgia, 1976 ..... . .........   68

  20   Estimated Production from New Cropland on Marginal, Submar-
         ginal, and Soil Bank Land in Georgia, 1976 .....................   72

  21   Estimated Increased Net Farm Income from New Cropland on
         Marginal, Submarginal, and Soil Bank Land in Georgia,
         1976 ........................................ ..... . .............   73

  Bl   Estimated Parameters for Interindustry Model of State
         Economies, 1975 .............. ... . .' ..............................   92
  Cl   Soil Resource Groups  (SRGs) of Georgia ................. ,

  C2   Total Hectares for Each Soil Resource Group and Land Resource
         Area in Georgia.
                                                                           99
  C3   Cultivatable, Marginal, Wetlands and Submarginal Lands in Row
         Crops by Land Resource Area ..... .
  C4   Cultivatable, Marginal, Wetlands and Submarginal Lands in
         Forest by Land Resource Area ...................................  104

  C5   Cultivatable, Marginal, Wetlands and Submarginal Lands in Other
         Uses by Land Resource Area ....... , .............................  105
  Dl   Runoff Curve Numbers  (CN) for Selected Agricultural Uses .........


-------
Number                                                                  Page

  D3   Plant Cover or Cropping Management Factor  (C) ..................   HO

  D4   Erosion-Control Practice Factor ................................   Ill

  D5   Estimated Sediment Delivery Ratio as Related to Average
         Drainage Area for Perennial Stream ...........................
  El   Estimated Runoff from New Cropland in Georgia, 1976 ............   115

  E2   Estimated Increased Sediment Yield from New Cropland in
         Georgia, 1976 ................................................   116

  E3   Estimated Sediment Yield from New Cropland in Georgia, 1976....   117

  E4   Estimated Nitrogen Loss from New Cropland in Georgia, 1976 .....

  E5   Estimated Phosphorus Loss from New Cropland in Georgia, 1976...

-------
                                   SECTION  1

                                 INTRODUCTION
     Farmers in the United States responded  to the  favorable profit situa-
tion for agricultural field crops during the 1973-1976 period by increasing
acreage of a number of commodities.  As a result, Soil Bank, marginal, and
submarginal land that was formerly not in cultivation is now being used for
crop production.  This increased agricultural production provided benefits
to the nation in increased domestic supplies and/or increased quantities of
agricultural commodities for export.  This increased production, however,
also has environmental effects, which are costs to  society that are not
accounted for in the cost-benefit calculations of farmers.  Increased agri-
cultural production results in additional fertilizers and pesticides that
may place additional pressure on the waste assimilative capacity of the eco-
system.  If the incremental land in crop cultivation is more subject to ero-
sion and runoff, the environmental impact of the increased production is
compounded.

     Past research by agricultural economists has typically not emphasized
the linkage between level of agricultural production and environmental
quality.  In fact, many aggregate studies of agricultural nonpoint source
pollution assume that crop acreage is constant (Wade and Heady, 1977; Alt,
Miranowski and Heady, 1979).  Even studies that allow adjustment of crop
acreages confine their analysis to reallocation of  crop acreages among vari-
ous crop alternatives, including hay.  In addition, these studies have not
considered alternative levels of demand for  agricultural commodities and
alternative government programs that caused  the recent increases in agricul-
tural production (Horner, 1975; Taylor and Frohberg, 1977; Osteen and Seitz,
1978).  Studies of agricultural price programs, which have been concerned
with the level of agricultural production, have likewise ignored the environ-
mental effects of level of agricultural production.  For example, a recent
comprehensive review of research on agricultural policy does not mention
environmental effects (Brandow, 1977).  Thus, the environmental impacts of
alternative levels of agricultural production have  been a relatively
neglected area of research.

     Georgia has many desirable features as  a case  study for measuring the
environmental impact of increased crop acreage.  Its land resources are made
up of a large number of soil series with widely divergent erosion potentials.
The land use configuration is balanced between forest, pastures, and culti-
vated crops.  In addition, the cultivated land is allocated to a large number
of crops relative to other states.  The combination of diverse land resources
and uses results in a variety of potential environmental loadings from

-------
 changes in land use.   Development of a methodology to analyze the environ-
 mental impact of increased crop acreage in Georgia, therefore,  would provide
 useful information for a number of situations.

      Such information will include geographic dispersions of increased
 amounts of pesticide  and fertilizer used and selected environmental loadings
 resulting from using  Soil Bank, marginal,  and submarginal lands for crop pro-
 duction.   While incorporating the environmental effects of increased crop
 production into a cost-benefit framework is beyond the scope of this study,
 public decision makers need this type of information to consider the full
 consequences of their decisions.  Information on environmental loadings from
 additional crop acreage may influence public policy decisions concerning
 levels of agricultural production as well as environmental standards for
 agriculture.  In particular,  such information should provide some insight
 into tradeoffs between environmental policies dealing with control of non-
 point source water pollution and agricultural production policies dealing
 with the land base used for crop production.
 OBJECTIVES

      The principal objective  of  this  report  is  to  present  a methodology to
 evaluate the environmental and economic  impact  of  placing  Soil Bank,  mar-
 ginal,  and submarginal land in crop production.  Specific  objectives  include:

      (1) Develop  a methodology that can  be used to:

           (a)  Estimate the recent  and potential changes  in land use by
      major land classifications  and soil associations;

           (b)  Identify crop yields and input usage, with particular atten-
      tion to fertilizers and  pesticides,  for selected row  crops by major
      land classifications  and soil associations;

           (c)  Estimate the environmental loadings  of soil,  pesticides, and
      plant nutrients in streams, lakes,  and  estuaries from various land
      uses of major  land classifications  and  soil associations;

           (d)  Assess the environmental impact of recent  changes in land
      use with  attention to net changes in environmental  loadings and  their
      impact  on dimensions  of  environmental quality;

           (e)  Assess the economic  impact of  recent land  use changes;  and

      (2)  Apply the methodology,  as a  case study, to the  state  of Georgia.


ORGANIZATION OF THE STUDY

      Section 2 includes conclusions from the study related  to  both its
methodology  and empirical  application.   Section  3  discusses and evaluates
the methodology used in this  study to estimate  the environmental and  economic

-------
impact of placing marginal, submarginal, and Soil Bank land in crop produc-
tion.  The remaining sections of the report deal with the empirical applica-
tion of the methodology to the state of Georgia.  Section 4 presents
information on Georgia land resources, including a geographic delineation
of the state by Land Resource Areas (LRAs).  Data presented in subsequent
sections are reported according to these LRAs.  Changes in land use patterns
and input usage for the period 1973-1976 are analyzed in Section 5.  That
analysis considers changes in row crops, pasture, and forest land.  Sections
6 and 7 analyze the environmental and economic impacts, respectively, of
placing marginal, submarginal and Soil Bank land in crop production in
Georgia for the 1973-1976 period.

-------
                                   SECTION 2

                            SUMMARY AND CONCLUSIONS
 METHODOLOGY

      Procedures were developed for delineating and characterizing land
 resources.   The basic unit of analysis  used in this study was the Soil
 Resource Group (SRG).  Physical properties  used to calculate erosion and
 runoff were determined for each SRG.  Soils were classified as suitable for
 cultivation, marginal, and submarginal  on the  basis of  the Soil Conservation
 Service land capability classification  system.   The advantage of this classi-
 fication system is that secondary data  are  readily available.  In order to
 focus on other land classification systems  that might account for such
 factors as  technical, aesthetic,  economic,  environmental and social variables,
 a completely new inventory of land resources would have to be taken.   Land
 identified  as Soil Bank land in this  study  was that land previously diverted
 from crop production under the Conservation Reserve Program (CRP)..  This
 category is of special interest because a large portion of this land could
 have been converted to crop production  in response to favorable farm prices
 experienced in the 1973-1976 period.

      Land use changes by SRGs and land  use  patterns were estimated from
 secondary data and surveys.   The  Conservation  Needs Inventory (CNI) (U.S.
 Department  of Agriculture, Soil Conservation Service, 1970)  provided bench-
 mark information on base land use patterns.  Changes in land use patterns
 were estimated from U.S.  Department of  Agriculture statistics on crop
 acreage,  surveys of  Agricultural  Stabilization and Conservation Service and
 Cooperative Extension Service,  and actual field surveys.   Changes in land  use
 on  Soil Bank land were estimated  using  survey  response  from participants in
 the CRP.

      Input-output levels  for major agricultural and forestry uses of  differ-
 ent  types of  land were synthesized from existing compilations of agronomic
 and  agricultural economic data for major soil  groups.   Representative input-
 output  levels in a budgetary framework  were developed.   Use  of fertilizers
 and  pesticides  for selected  crops on  various soils in Georgia was of  parti-
 cular interest  in considering potential environmental impacts.   In addition,
 these soils were characterized according to factors affecting erosion.   After
 land use changes  were determined,  quantities of  fertilizers  and pesticides
 used were estimated.

     Major attention was  focused  on estimating  the net  changes in environ-
mental  loadings of sediment,  pesticides, and fertilizers.  Sediment yield

-------
for each SRG was based on estimates of sheet and rill erosion by applying the
universal soil loss equation  (USLE) and sediment delivery ratios based on
size of drainage area and soil texture.  Estimated runoff, estimated sediment
yield and fertilizer application rates were used to determine nitrogen and
phosphorus losses.  Evaluation of pesticides was based on an index of poten-
tial environmental impact.  The relative impacts of environmental loadings on
various dimensions of environmental quality were assessed only at the con-
ceptual level.

     It is extremely difficult to thoroughly analyze all the environmental
costs that may result from placing forest and pasture land in crop production.
The ubiquitous nature of the  externalities that result from increased environ-
mental loadings contribute to the difficulty of conducting a complete, tho-
rough analysis.  Data needs for such a broad study would be immense.  Even if
many of these costs were accounted for and measured, political decisions that
consider the distribution of  impacts would be necessary to compare benefits
of increased crop production with environmental costs.  Consequently, environ-
mental loadings are estimated in this study rather than environmental costs.

     The net economic benefits of changes in land use patterns were assessed
with particular attention given to estimating the incremental value of agri-
cultural production, net farm income, and the secondary impact on the economy.
Procedures were developed to  estimate production and net farm income for
selected crops, pastures and  forests.  An econometric interindustry model
was developed and used to estimate the effect of a change in agricultural
production on aggregate personal income and employment.

     This methodology could be utilized in similar studies in other areas of
the United States.  Most of the data required in the study is either collected
in all areas of the country or should be available as an adjunct to normal
state agricultural research and extension.  The major advantage of the
methodology is that the data are largely secondary so that time and resources
required for data collection  are minimized.  The methodology also has the
advantage of being linked to  data which represents actual behavior on land
use, crop yields, pesticide and fertilizer use.  In this study, only state
wide data were available on a number of economic variables which were utilized
to estimate net farm income and input use.  In some states, more sub-state
data would be available, which would improve the accuracy of components of
the research.  The major weaknesses of the methodology relate to the large
number of assumptions necessary to implement the procedures without survey
data.  Unfortunately, data were unavailable to validate most of these assump-
tions in this study.  Before  the methodology is employed on a wide scale, a
follow-up study would be advisable to validate these assumptions.  Such a
study would contrast results from general surveys with results from the
assumptions involved in this methodology.  The primary data necessary to
validate the methodology are discussed in Section 3.  Such a future study
would be useful in delineating which assumptions are a good approximation
and which areas of the methodology should be supplemented with primary survey
data.

-------
      It must be stressed  that this methodology has been designed  to provide
 information for policymakers on the nature and approximate magnitude  of  eco-
 nomic and  environmental variables associated with land use changes at an ag-
 gregate level.  Many  of the possible biases in the methodology at the micro
 level probably counter affect each other at the macro level.  If  similar
 information was required  for a planning model either at the state or  sub-
 state level, the methodology must be validated in a follow-up study to pro-
 vide  confidence for analytical purposes.
 EMPIRICAL  APPLICATION

      This  study  has  attempted to quantitatively measure the environmental  and
 economic impacts of  placing marginal, submarginal, and Soil Bank  land  in crop
 production in  Georgia.   It was estimated that  74.5 thousand hectares of such
 land  was placed  in crop  production in Georgia  during the  1973-1976 period.
 The application  of the methodology indicates that substantial environmental
 loadings of sediment, pesticides, and fertilizers would result from new crop
 production on  marginal,  submarginal, and Soil  Bank land.  The pollutants
 generated  annually from  each hectare of this land in crop production,  on the
 average, are estimated to be 11.9 metric tons  of sediment, 8.1 kilograms of
 nitrogen loss  (excluding leaching), and 10.5 kilograms of phosphorus loss.
 However, considerable variation exists among subclasses of these  lands.  For
 example, the increased annual sediment yield per hectare  from this new crop-
 land  ranged from 0.74 to 32.41 metric tons in  different LRAs.

      Increased pollution also varied between marginal and submarginal  land
 and cultivatable Soil Bank land:  the state average for loss of nitrogen per
 hectare was 7.16 kilograms on marginal and submarginal lands and  9.13  kilo-
 grams on cultivatable Soil Bank land.  The higher average for cultivatable
 Soil  Bank  land resulted  from the relative concentration of new cropland in
 LRAs  more  subject to nitrogen loss; for most LRAs, the nitrogen loss was
 smaller for cultivatable Soil Bank land than for marginal and submarginal
 land.  The  same  difference between these two categories also existed for sedi-
 ment  and phosphorus  loss.  These results indicate that the conventional mar-
 ginal and  submarginal classification is not necessarily a good index of
 pollution  from crop  production.  For control of agricultural pollution, exami-
 nation of  broader dimensions than conventional soil conservation  classifica-
 tion  will be necessary.

      Erosion appears to  be a key factor that affects environmental loadings of
 pollutants  from  cropland converted from marginal, submarginal, and Soil Bank
 land  in Georgia.  First, the magnitude of erosion from crop production activi-
 ties  on marginal, submarginal, and Soil Bank land is generally higher  than on
 land  classified  as suitable for cultivation.   Secondly, soil loss from crop-
 land may carry with  it nitrogen, phosphorus, and pesticides.  Chemical loss
 could  be reduced considerably in many cases with a reduction in erosion.

      Even without being  bound to soil particles, pesticides and nitrogen can
move  away from application sites through leaching and runoff.  In Georgia,
runoff and  leaching  is only slightly higher on marginal, submarginal and
 Soil Bank land than  on land classified as suitable for cultivation.

-------
     Increasing crop acreage  (thus, prop production) has substantial benefi-
cial economic impacts that must be weighed against increased environmental
loadings.  The 74,5 thousand hectares of marginal, submarginal, and Soil Bank
land placed in crop production during the 1973-1976 period increased crop
production $23,6 million annually.  Net farm income increased $4.5 million
in aggregate or $61 per hectare annually.  Increased crop production has a
multiple impact on Georgia's economy.  Agribusiness would supply more inputs
to farmers and process and handle more agricultural products.  In addition,
farmers and agribusiness workers, as consumers, would purchase goods and
services from numerous sectors of the economy.  Assuming that this increased
level of production and farm income would be sustained, aggregate personal
income would increase $9.1 million annually.

     The new cropland placed in production in 1973-1976 represents only a
small fraction of total marginal, submarginal, and Soil Bank land that could
be placed in crop production in Georgia in the future.  Over 5 million hec-
tares of marginal and submarginal land in Georgia has a potential for crop
production.  Land identified as having a potential for crop production ex-
cludes very steep and stony land but includes land that may require some con-
servation measures before it could be placed in crop production.  The Piedmont,
Southern Coastal Plain, and Atlantic Coast Flatwoods each contain over 1
million hectares of this land.  However, the other LRAs also contain sub-
stantial acreages of marginal and submarginal land that could be used for
crop production.

     Most of the Soil Bank land that had not been placed in crop production by
1976 remained in trees.  It was estimated, however, that from 12.5 to 14.5
percent of the acreage in pine plantations on Soil Bank land had been cleared
by 1976.  Because trees on Soil Bank land are reaching sawtimber size and are
producing $74 worth of wood (using 1976 prices) per hectare per year on the
average, little of this land is expected to be cleared in the near future.
In the long-run, trees on this land will be harvested, and if crop prices
are favorable this land could eventually be placed in crop production.

-------
                                   SECTION 3

                                  METHODOLOGY
      The methodology developed in this  study to  estimate ex post environ-
 mental and economic impacts of land  use changes  is  presented and evaluated in
 this section.   For ease in exposition,  the methodology  can be segmented into
 several major  components as shown in Figure 1.   The first segment deals with
 inventory of land resources within the  area under consideration.   In this
 segment, basic soil resources and their potential capability for agricultural
 use are defined.   Marginal, submarginal,  and Soil Bank  land are also defined.
 The first segment is primarily descriptive but provides concepts utilized in
 later analysis.

      The first analytical component  involves estimation of land use changes.
 Both the magnitude and the form of changes,  as well as  identification of soil
 resources on which the land use changes occurred, are considered in this
 component.  The next component involves estimation  of changes in agricultural
 input use associated with changes in land use from  pasture and forestry to
 crops.   Here emphasis is placed on fertilizer and pesticide use,  both of
 which are important in considering environmental as well as economic impacts.
 The final two  methodological components involve  estimation of the environ-
 mental and economic impacts of land  use and associated  input changes.   Envi-
 ronmental impacts considered include loadings of sediment,  nitrogen,, phos-
 phorus,  and pesticides.   The economic impacts focus on  two components:
 (1)  changes in farm income, and (2)  the impact of changes in farm inputs and
 income on other sections of the economy.   Estimation of the external (i.e.,
 societal)  costs  of changes in environmental loadings was beyond the scope of
 this study.  The  nature of these costs  and problems associated with their
 measurement, however,  are summarized in the economic impact section.
LAND RESOURCE  INVENTORY

     The methodology used  in  inventorying  land  resources  is  summarized in
Figure 2.  The most aggregate level  is  Land  Resource Regions.   The continen-
tal United States is divided  into  20 Land  Resource  Regions,  based primarily
on land uses within each region.   Each  region is  subdivided  into Land Resource
Areas (LRAs), which are more  representative  of  general  classes  of soil found
within each Resource Region.  Much of the  information in  this report  was  col-
lected at the county level and compiled by LRA.   Estimation  of  the environ-
mental and economic impacts associated  with  land  use configurations and land
use changes requires more detailed knowledge of certain chemical and  physical
properties related to individual soil series  or  groups of  soil series  called
associations.

                                       8

-------
               Estimate Land
                Use Changes
              Estimate Changes
              in Agricultural
                  Inputs
   Estimate
Environmental
    Impact
Estimate
Economic
 Impact
   Figure  1.   Overview of methodology.

-------
                                             Land Area
                                         Land Resource Areas
                                        	(LRA)	
                                   Counties and Percent of Counties
                                        Within Individual LRA
                                        Breakdown of Counties
                                       into Soil Bank Land and
                                 All Land Which is Suitable, Marginal
                                 and Submarginal for Crop Production
By SCS Capability Classes from
the Conservation Needs Inventory
Two Methods Us
 for  Comparlso
;ed \
m  )
By SCS Capability Classes Established
	for  Soil Resource Groups	

Le
t


Suitable
for Crop
Productior
(I-III)

|
ind Under Conservation
se (Soil Bank Land)





Marginal
for Crop
Production
(IV)



Submarginal Sui
for Crop for
Production Prod
(V-VIII) (I-


Wet lands
(V)
1

1 All Other Land
1 in the County





table Marginal
Crop for Crop
action Production
III) (IV)





Submarginal
for Crop
Production
(V-VIII)


1 Wetlands
| (V)
   Figure  2.   Schematic  showing  how land  resources were inventoried.

-------
     Areas representing soil association were planimetered from a soil asso-
ciation map.  First, individual county areas were planimetered.  Then soil
associations within each county were planimetered.  Planimetering was judged
to be at least 98 percent accurate when comparing derived county land areas
with other sources.  These data were used to calculate the area of various
soil associations within each county.  Slope ranges were obtained from the
soil association map, which included ranges of 0-5%, 5-12%, 12-25%, 25-60%,
and >60%.  The most dominant soil series found within each soil association
was used to establish Soil Resource Groups  (SRGs).  Soil associations were
given the same name as the dominant soil series and were considered to have
properties similar to that series.  Different associations dominated by the
same series were given the same SRG name.   Some SRGs may have the same
general name, but are different because of  their slope.  It should be noted
that grouping soil series together and assigning the characteristics of the
dominant soil series to the entire group is an oversimplification that may
bias the results.  The degree of bias is an empirical question that will be
considered in the empirical section of this report.  Disaggregating the
analysis to the soil series level would have greatly compounded the amount
of soils data and would have increased problems in establishing correspon-
dence of soils data to other secondary county data.

     To estimate chemical and sediment loadings, each SRG was characterized
by those physical properties important in calculating erosion and runoff.
Dominant agricultural production capability classes and subclasses were
established for each SRG, based on the average slope associated with dominant
soil series.

     Marginal and submarginal lands, as used in this study, are based on the
Soil Conservation Service sys-tem of land capability classification.  Under
this system, land is classified according to its ability to produce culti-
vated crops and pasture without deterioration over a long period of time
(Hockensmith, 1948).  The capability system groups soils by common management
problems.  Classes I-III are classified as  suitable for cultivation.  Class
IV is marginal, Class V is wetlands, and Classes VI-VIII are submarginal for
crop production.  Classes V-VIII are grouped and classified as submarginal
lands for purposes of this study.

     The land area in each county classified as suitable for cultivation,
marginal, and submarginal for crop production, as established for SRGs, were
compared (see Figure 2) with published data on capability classes reported in
the Conservation Needs Inventory, 1967 (U.  S. Department of Agriculture, Soil
Conservation Service, 1970).  This comparison provides some measure of
accuracy of the approach used in this report.

     The Soil Bank program was established by the Agricultural Act of 1956 in
an attempt to reduce agricultural production and thus bring production in
line with utilization (.U.S. Department of Agriculture, Agricultural Stabili-
zation and Conservation Service, Farm Commodity and Related Programs, 1976).
This Act contained two principal provisions:  (1) under the Acreage Reserve
provision, farmers were paid to reduce planted acreage of allotment crops.
The Acreage Reserve was operated under annual agreements with producers for
three years; and (2) under the Conservation Reserve provision, farmers were

                                       11

-------
paid  to divert all  or part  of  their  cropland  to  soil-conserving  uses  under
long-term  contracts.  The Conservation Reserve Program was  a  general  cropland
retirement program.

      To participate in  the  CRP farmers contracted  to withdraw a  designated
area  for a period of three  to  ten years.  A conservation  cover (trees or
grasses) had  to be  maintained  on the diverted land.  In addition,  each par-
ticipant's cropland acreage had to be reduced by the amount of acreage
placed in  the CRP.  For participating in the  program, farmers received an
annual rental payment to compensate  for the loss of income  that  the acreage
otherwise  would have produced.

      The CNI  of 1967  (U.S.  Department of Agriculture, Soil  Conservation Ser-
vice, 1970) was used to identify Conservation Reserve acreage by soil capa-
bility classes and  subclasses.  That publication reported acreage  of  land
under the  heading of "conservation use only"  in  1967.  Although  this  cate-
gory  has a broader  definition  than Conservation  Reserve,  a majority of the
acreage was Conservation Reserve acreage.  The acreage restricted  to  "con-
servation  use only" will be referred to in subsequent sections of  this report
under the  general term  of Soil Bank  land.

      One of three alternative  units  of analysis  could have been  used  in the
study—hydrologic units, water quality management  units,  or political units.
Hydrologic units, which designate drainage areas with all water  in the area
draining to a selected  point,  could  have been the  primary unit for analysis.
A second alternative could  have been the selection of water quality manage-
ment  units, which are generally bounded by rivers  or smaller  streams  and
ridge tops.   Finally, political units such as counties or planning areas
could be used as the unit of analysis.  Soils data can be planimetered on  the
basis of any  of these units.   Secondary sources  of land use are  available  for
counties but  not for hydrologic or water quality management units.  Economic
data  related  to agricultural production and input  usage would only be avail-
able  at the county  level.   Also, data on land in the Soil Bank are available
only  at the county  level.

      Land  was classified in this study according to its ability  to produce
crops and  pasture.  This avoids the  very serious difficulties in defining
land  types that have recently  arisen in the national debate on prime  agri-
cultural lands (U.S. Department of Agriculture,  Perspectives  on  Prime Lands,
1975).  Preservation of prime  agricultural lands has been a popular cry, but
the scientific community has been unable to clearly delineate criteria to
define prime agricultural land.  The problem  is  complicated by the fact that
an appropriate classification  system for prime agricultural land would have
to include technical, aesthetic, economic, environmental, and social  vari-
ables.

     The approach used  relies  on soil classification criteria developed by
soil scientists.   The advantages of  this approach  are twofold.   First, the
classification system is well  understood because it is widely used.   Secondly,
land resources in most  states  have already been  classified with  this  system in
the CNI  of  1967 (U.S.  Department of Agriculture, Soil Conservation Service,
                                      12

-------
1970).  Any other approach involving  a  complete  inventory  of  the state's  land
resources would be beyond the scope of  this  study.


LAND USE CHANGES

     The overall process for estimating  land use changes is shown in Figure 3.
Land use changes are calculated at the  county level  and aggregated for LRAs
and for the state as a whole.  Net additions to  crop acreage  for the period
1973-1976 were determined from annual estimates  by the Economics, Statistics,
and Cooperatives Service (Georgia Agricultural Facts, 1978).  Although total
crop acreage in each county is not reported  directly, it can  be estimated
using acreages of the major crops.  Only those counties that  experienced an
increase in crop acreage during 1973-1976 were assumed to  bring new cropland
into production.  Some new cropland may  have been brought  into production in
other counties, but such counties would  experience no net  addition to crop
acreage.

     Increased crop acreage may involve  the  conversion of  several land-use
patterns to crop production.  For example, woodland, pasture, wildlife areas,
idle lands, or wetlands may be converted to  crops.  Woodland  and pasture are
particularly important for economic considerations because these lands
generate marketable products and farm income that must be  foregone to produce
crops.  Surveys conducted by the Soil Conservation Service and Cooperative
Extension Service were used to identify  types of land use  conversion to crop
production in each county.

     Having determined general row crop  changes  for each county, the next
step involved relating crop changes to  specific  soil series or soil associa-
tions.  Two approaches were used to identify increases in  crop acreage by
SRG.  LANDSAT remote sensing technology  was  used in the first approach.
LANDSAT scenes that contained a large portion of agricultural land were chosen
for analysis.  A sample of counties within each  of the scenes was identified
on the basis of soils and crops as representative of the particular LRAs.
Land use- changes within a county were superimposed on county soil maps in
order to quantify land use changes according  to  SRGs.  A more detailed de-
scription of the LANDSAT procedure is presented  in Appendix A.

     The second approach used to link cropland changes with SRGs relied on
the CNI of 1967 (U.S. Department of Agriculture,  Soil Conservation Service,
1970).  New cropland was assumed to be distributed among the classifications
of suitable, marginal, and submarginal  land  in the same proportions as land
already in cultivation as reported in the CNI of  1967 (U.S. Department of
Agriculture, Soil Conservation Service,  1970).   Within the classifications
of suitable, marginal, and submarginal  land  for  cultivation, new cropland
was allocated to the various SRGs in  proportion  to their relative shares of
the land area.  For example, if suitable (Class  I-III) land in a county was
in two SRGs with the same number of acres then new cropland on this land was
assumed to be equally distributed between these  two SRGs.
                                       13

-------
                                                            Major Land Use  Changes
                                                            	to Cropland	
                              Counties Showing Increased Row
                                 Crop Area from 1973-1976
From:
1.  ASCS  Reports
2.  Forest  Service
   Inventory
3.  Survey  of
   Participants
                                                                                 Counties Showing No Increased
                                                                                  Row  Crop Area from 1973-1976
                                               Pasture or Grassland
                                                       to
                                               	Cropland	
                                                   Cropland to Other  Uses
                                                    Including Urban Was
                                                      Not Considered
From:
1.  Forest  Service
   Reports
2.  SCS Survey
3.  County  Extension
   Survey
4.  Landsat Remote
   Sensing	
From:
1.  County Extension
   Survey
2.  SCS  Survey
3.  Landsat Remote
   Sensing
                      Figure  3.   Schematic  showing identification of land  use changes.

-------
     The economic theory of land use provides a basis for assessing some of
the methodological assumptions utilized in this study.  The neoclassical
theory of land use suggests that landowners select  the appropriate use for
land on the basis of the sum of discounted future net revenues from alterna-
tive land uses (Barlowe, 1978; Alonso, 1964).  From this viewpoint, land
would shift from pasture and forestry to cropland whenever the perceived sum
of discounted future returns from crops became greater than from other uses.
This basic hypothesis has some important implications for the methodology in
this study.  First, one would not necessarily expect all land of similar
quality to be shifted at the same time.  Because these decisions are based
on the subjective expectations of future revenues,  the expectations of dif-
ferent landowners would likely be different.  In addition, the stage of
maturity of the timber on some of the forestland may provide incentives for
delayed harvesting.  Secondly, shifts in land use would not necessarily be
limited to one type or quality of land.  Although land more suitable for
cultivation might have higher returns from crop production than marginal
or submarginal land, increased crop acreage would not be limited to this
classification of land.  Marginal and submarginal lands may have higher
returns in crop production than in other uses.  Landowners, therefore, would
also shift some marginal and submarginal land into  crop production if they
have expectations that crop production is more profitable than other uses.

     The first methodological approach discussed in this section of using
survey information to determine the land us.e patterns would be the pre-
ferable approach.  In the absence of sufficient resources to support a
large scale empirical study, however, the second approach does have merit.
Basing the proportion of new crop land that is suitable, marginal, and
submarginal on historical proportions takes into consideration previous
economic decisions related to appropriate land use.  This approach requires
the assumption that land use changes were based on  similar economic calcu-
lations as those associated with historical patterns.  Although this assump-
tion can be questioned, it does have an empirical basis.  Research to vali-
date this assumption would be useful, but without these data the second
approach- utilized here can be used in land use and  environmental research.
CONVERSION OF SOIL BANK LAND

     Most of the land that was placed in the CEP  (Soil Bank program) was pre-
viously used for crops, hay or pasture.  Hence, it was hypothesized that a
significant portion of this land would be converted from conserving uses to
crops in the 1973-1976 period.  Determination of  the status of this land was
based on surveys of participants in the Soil Bank program.  The list of Soil
Bank participants was stratified by geographic areas of the state to assure
that selected samples would be geographically representative.  Then 10 percent
of the participants in each stratum were randomly selected and surveyed
through mail questionnaires.  A follow-up questionnaire was sent to ensure
adequate response.
                                       15

-------
 INPUT USAGE

      The type and quantity of inputs was estimated for cropland being brought
 into production.   The use of chemical fertilizers and pesticides is of pri-
 mary importance because these inputs may adversely affect public water sup-
 plies, recreation opportunities,  as well as contribute to the cost of crop
 production and, therefore, may result in both direct and off-site economic
 impacts.  Two methodological problems are involved in estimating chemical
 input use.  The first problem involves the paucity of secondary data for
 quantities of agricultural inputs.   Unlike crop acreages and production esti-
 mates, chemical use is not routinely derived.   The problem is especially acute
 for pesticides, for which the only  general source of information is periodic
 surveys by the U.S. Department of Agriculture.

      The second problem involves  obtaining sufficiently disaggregated data on
 input use.  Many secondary input  use data are state estimates,  and, in the
 case of pesticides, the data are  for multistate regions.   For this study the
 problem was crucial, because economic theory would lead one to expect that
 input use would vary spatially, especially among SRGs.   Because the marginal
 productivity of an input could vary among different SRGs,  the amount of
 input used would likely vary.   This proposition has special validity for
 fertilizer use, where differences in natural fertility and capacity of the
 soil to retain fertilizer nutrients throughout  the production season would
 lead to different crop responses  from fertilizer.   In similar manner, dif-
 ferences in soil and climate among  different regions could lead to differences
 in incidence of weeds, insects, and fungi so that the productivity of pesti-
 cides could also vary.

      These economic hypotheses suggest that the appropriate data for estima-
 tion of input use would have to be  collected through surveys of the area
 under consideration.   If county data on input use are unavailable,  an alter-
 native approach would be to apply regression analysis to  estimate use on
 particular soil groups that are larger than counties.   According to Ziemer,
 Musser,  and Clifton (1978),  this  approach involves estimating the following
 equation using dummy variables.


                  ^  = I \ °ik + E al Yil + E  am Dik Yil                 
-------
third approach. Is to utilize the expert knowledge of researchers and exten-
sion personnel in the area under consideration  to synthesize estimates of
input use.  With sufficient breadth of specialized experience, it would be
possible to synthesize estimates for various soil provinces or other sub-
regions under consideration.  This methodology  has been extensively used in
production economics and farm management research as well as in previously
referenced economic studies of agriculture related nonpoint pollution.

     Because of limited data, the third approach was utilized in this study. .
Use data were developed for each crop and for the state as a whole.  State
rather than substate data were used for two reasons.  First, previous research
with aggregated input data indicates no significant differences in input use
among provinces in the state (Ziemer, Musser and Clifton, 1978); and second,
there is a lack of sufficient expertise to synthesize substate estimates.
These estimates were designed to be representative of use rates for all
levels of management in the state.  For fertilizer inputs, previously syn-
thesized estimates were modified and reviewed by personnel in the Georgia
Cooperative Extension Service and the Economics, Statistics, and Cooperatives
Service, USDA.

     For pesticides, representative use rates were not available on a state
level for the period of analysis.  The procedure used to estimate pesticide
use involved updating the Pesticide and General Farm Survey that was con-
ducted in 1971  (USDA, Economic Research Service, 1971).  Lists of chemicals
identified in that survey along with amounts used per hectare in 1971 were
sent to state specialists in the Cooperative Extension Service to estimate
the type and amount of chemicals used in 1976.  These specialists were asked
to estimate amounts and levels of chemicals in  use in 1976 that would have
replaced chemicals no longer used.
ENVIRONMENTAL IMPACTS

     The nonpoint agricultural pollutants of primary concern are sediment,
nitrogen, phosphorus, and pesticides.  Pesticides, fertilizer, and sediment
are delivered to water bodies primarily by water runoff.  Runoff transports
dissolved water-soluble  chemicals, as well as chemicals bonded to sediment
particles.  Runoff can originate from both rainfall and irrigation; however
in this study, the analysis focused on rainfall.  Figure 4 shows factors that
were considered in evaluating the environmental impact associated with
increased crop acreage.

Sediment

     Sediment affects the quality of water resources directly but also may
carry pesticides and plant nutrients adsorbed on the soil particles.  The con-
tribution of agricultural land to sediment in waterbodies is dependent pri-
marily on the erosion rate and the sediment delivery ratio, i.e. the propor-
tion of eroded soil that reaches the stream.  Using representative soil,
cover, and topographic features, it is possible to estimate the erosion rate
for relatively homogeneous agricultural areas.
                                       17

-------
                                                                            State Level
                                                                       | Land Resource Areas
                                                                           |  Counties
                                                                      Specific Land Use Changes
                                                                    Were Adjusted to Individual
                                                                       Soil Resource Groups	
                                       r
                       Land Use Changes on Soil Resource
                   Groups  Classified as Soil Bank, Marginal
                   and Submarginal Land for Crop Production
                           Land Use Changes on Soil  Resource
                           Groups Classified as Suitable for
                          	Crop Production	
                                                                                                                        No Further Calculations
oo
                                        Inventory of Pesticide Use to
                                        Include:
                                        1. Toxicity
                                        2. Persistence
                                        3. Biomagnification
                                        4. Method and Time of Application
                                        5. Transport Mode
                                        All factors can be compared to
                                        estimate potential environmental
                                        hazard
                                                                                                  |  Economic Impacts
Farm Impacts
  Production Inputs
  Crop Production
  Net Farm Income
Costs to Society
 from Sediment
   and Other
   Pollutants
Aggregate Economic
	Impact	
                             Figure  4.   Schematic  showing evaluation of  economic and environmental
                                               impacts  associated with increased cropland.

-------
     Sediment yield per hectare  of  agricultural  land, which  depends on  estima-
tion of potential soil loss utilizing  the Universal  Soil Loss Equation
(Wischmeier and Smith, 1965)  and sediment delivery ratio, may be expressed as:

                           SY  =  (R-K-L-S-OP)  S,                           (2)

where

     SY is sediment yield  from surface erosion in tons per hectare

     R is the rainfall and erosivity index

     K is the soil-credibility factor  which depends  on soil  properties

     L is the slope length

     S is the slope steepness

     C is the cover and management  factor

     P is erosion control  factor, with no control P  = 1.0

     S  is the sediment delivery ratio.

The factors R, K, L, and S, which are  fixed for  a given location under exist-
ing management, determine  the location's characteristic erosion potential.
Factors C and P, on the other hand, are determined by management.  The cover
and management factor C is related  to  the soil's ability to  resist
erosive forces  as affected by vegetation cover, crop sequence, management,
and agronomic erosion control factors.  Factor P concerns such cultural
practices as contour plowing  and terracing.

     Soil topographic details such  as  soil properties, slope steepness, and
slope lengths were determined for each SRG in the state.  Soil survey maps
from the Soil Conservation Service  were the primary  sources for soil charac-
teristics and associated land features.  These maps  included classifications
of erosion potential and land slope.   Land slope class indicates whether the
land is nearly level (0-5  percent slope), gently sloping (5-12 percent),
moderately sloping (12-25  percent), strongly sloping, or steep-  Typical
cultural practices were identified  through numerous  field inspections and
interviews with local extension,  ASCS,  and SCS workers.

     The amount of sediment that actually reaches a  designated point such as
a pond, a reservoir, or a  point  on  a stream or river is called the sediment
yield.   The soil loss equation computes long-term gross sheet and rill
erosion but does not directly predict  downstream sediment yield.  The dif-
ference in the amount eroded  from a total drainage area and  that transported
to a selected point determines the  delivery ratio.  Many factors influence
the sediment delivery ratio.  No general equation for watershed delivery
ratios  has been derived, but  several observed relationships provide guide-
lines for approximating them.  Size of  drainage area is an important factor,


                                      19

-------
 because  as  distance  of  sediment  transport  to downstream points  increases
 opportunities  for  deposition  enroute are more numerous.   Sediment  yield and,
 therefore,  the delivery ratio decrease.  Sediment delivery ratios  vary widely
 for  any  given  size of drainage area, but limited data  show that, roughly,
 these  ratios are inversely related  to  size of drainage area  and directly
 related  to  surface texture of the more dominant soils  within the watershed
 (Roehl,  1962).

     One enigma of determining sediment delivered with this  procedure is the
 decision of what size drainage area to use in making calculations.   Obviously,
 the  larger  the stream the larger the drainage area and thus,  the smaller the
 sediment delivery  ratio.  The sediment yield, therefore,  can be artificially
 reduced  by  considering  only larger  watersheds as the appropriate drainage
 area.  Given the state  of research  on  estimating and assigning  dollar values
 to social (or  off-site) damages  of  sediment, the selection of the  size of
 drainage area  may  be somewhat arbitrary.   In some cases the  selection will
 be dictated by the nature of  the study (e.g., desired  scale  of  analysis),
 jurisdictional boundaries (e.g., 208 planning areas),  or  resource  planning
 areas.   In  this study,  the smallest drainage areas delineated for  LRAs were
 used.  Although this methodology may overstate sediment of public  concern, it
 was  decided to give  maximal rather  than minimal estimates.

 Runoff

     Runoff, which is precipitation that appears in surface  streams, occurs
 during and  shortly after a storm.   The two categories  of major  interest in
 this study  are surface  runoff and subsurface runoff or interflow.  Surface
 runoff reaches the stream by  traveling over land, whereas subsurface runoff
 infiltrates the surface soil  before reaching the stream.  It is practically
 impossible  to  separate  surface and  subsurface runoff,  so  these  are combined
 for  analytical purposes here  and called direct runoff.  For  agricultural
 land,  surface  runoff is generally the  major component  of  direct runoff, both
 in terms of quantity generated and  pollution potential.   Some areas with tile
 drains,  for example, may be exceptions.

     The amount of runoff from agricultural land must  be  estimated in order
 to measure  fertilizer and pesticide losses.  Surface runoff may carry fer-
 tilizers and pesticides in solution, in suspension, or adsorbed to suspended
 soil particles.  Subsurface water can  carry soluble chemicals.

     The amount of runoff from agricultural land due to a storm event is
 related  to  a host  of soil parameters and types of vegetation.   Direct runoff
 is estimated here  for specific storms  using empirical  procedures developed by
 the Soil Conservation Service, USDA.   SRGs were assigned  hydrologic  charac-
 teristics of the most dominant soil series.  These hydrologic characteristics
 influence the amount of  infiltration for bare soil.  The  effect of land use
 and management practices on runoff  has also been determined  by  the Soil Con-
 servation Service.   The  effect of combined hydrologic/soil-cover characteris-
 tics on  the amount of runoff  is  represented by a runoff curve number (CN).
Potential abstraction (S), that portion of precipitation  that does not con-
 tribute  to direct  runoff, is  related to soil and cover conditions  of a water-
shed as  follows.

                                      20

-------
Abstraction  and  storm rainfall are considered in measuring direct runoff  as
follows.

                                   CRf.  - 0.2 S)2

                              Qi = Bf^+O.BS                            C4)
                                     i
where:

     Q. is direct  runoff for ith storm  event

     Rf.  is  total  storm rainfall

     S is potential abstraction

     Annual  potential direct runoff was based on the  frequency and amount of
runoff for 1,  2, 5,  10,  25,  50, and 100 year storms.   Total storm rainfall
 (M±) is  measured  over a 24-hour period.   The expression  for annual potential
direct runoff  is as follows.


                               AQ -  Z  ^ Q.                              (5)
                                    i-1   ±

where:

     AQ is annual  potential  direct runoff

     Q. is direct  runoff for ith storm

     Y. is 1,  2, 5,  10,  25,  50 or 100 year storm

Hence, AQ is an expected value—the sum of each  runoff multiplied by the
probability  of that storm occurring.  This formulation assumes that in any
year, all of the land will receive a one-year frequency storm, half of the
land will receive  a two-year frequency  storm,  etc.  Fertilizer loss is
expected  to  be directly  related to the  estimate  of  annual  potential direct
runoff, but  pesticide loss depends more on the distribution of direct runoff
within a  year.

Fertilizer Loss

     Chemical fertilizers applied to  cropland may reach waterbodies through
leaching, runoff,  and  erosion.   The two nutrients of  environmental concern
are nitrogen and phosphorus.   Studies in which nitrogen and phosphorus losses
from farmland were measured  have found  a positive correlation between land
use and nitrogen and  phosphorus levels  in surface water (Baker, Johnson,
Borcherding, and Payne,  1979;  Holt,  1969;  Jaworski  and Hetling, 1970; and
McCarl, 1971).  An analysis  by Omernik  indicated that in  the eastern United
                                       21

-------
States streams draining agricultural lands had five-fold higher total nitro-
gen concentrations than those from forested watersheds.

     The behavior of nitrogen in a soil-plant system is complex.  Fertilizer
nitrogen is taken up by plants, retained in the soil organic fraction,  lost
by denitrification, and lost by leaching, runoff, and erosion.  The Committee
on Nitrate Accumulation, National Academy of Sciences estimated that 10%  to
15% of fertilizer nitrogen is lost to waterways.  Of this total, 49 percent
is transported by erosion, 32.7 percent by leaching of native soil nitrogen,
and 18 percent by transport of fertilizers by a combination of processes.
Phosphorus is not lost from the soil as rapidly as nitrogen, but remains
fixed and slowly becomes available for crop growth.  The predominant mode of
transport for phosphorus is soil erosion (McElroy, et^ al_. , 1976, pp. 102-106).
Total phosphorus losses for the United States have been estimated at about 25
percent of the phosphorus made available annually in the form of fertilizer
and livestock wastes (Wadleigh, 1968).

     Only a limited number of scientific approaches have been used to estimate
fertilizer contributions to nutrient contamination of water (National Research
Council, Nitrates:  An Environmental Assessment, 1978).  Controlled studies of
watersheds are costly, and the data provided are site-specific.  Smaller-scale
field experiments have limited applicability for extrapolation to larger
areas.  Use of tracer materials is not feasible because of the high cost.

     The heterogeneity of soils and of hydrologic characteristics from site
to site make general quantitative relationships for nitrogen and phosphorus
losses difficult to define (National Research Council, Nitrates;  An Environ-
mental Assessment, 1978).  The magnitude of losses of these two nutrients
from chemical fertilizers are calculated in this study by analyzing fertilizer
input usage and mode of transport to surface water.  In this study it is
assumed that nitrogen loss is directly linked to sediment and runoff rates
through the following relationship.

                                      SY.      AQ.
                        NL. = AN. [NS ~^- + NR -^ j                       (6)
                          1     x     SY       AQ

where

     NL. is nitrogen loss per hectare of cropland for ith SRG

     AN. is application rate for nitrogen for ith SRG

     NS is percent of applied nitrogen lost with sediment, on the average

     SY. is sediment yield for ith SRG

     SY is state average sediment yield

     NR is percent of applied nitrogen lost with runoff, on the average

     AQ.  is annual potential direct runoff for ith SRG


                                      22

-------
     AQ is state average potential  direct  runoff

According to this relationship, nitrogen loss  due  to  erosion  on  land with
potential sediment yield that  differed  from the state average would differ
proportionately from  the average  amount linked to  erosion.  Also, nitrogen
loss on land with a potential  runoff  rate  that differed from  the state
average would differ  proportionately  from  the  average amount  linked to runoff.
Using a similar procedure, phosphorus loss  is  directly linked in this study
to the fertilizer application  rate  and  sediment as the predominant transport
mode.

                                             SY
                               PLi = **± [PS ^T 1                       C7)

where

     PL^ is phosphorus  loss per hectare of  cropland for ith SRG

     AP. is application rate for  phosphorus  for ith SRG

     PS is percent of applied  phosphorus lost  with sediment,  on the average

     SY± is sediment yield for ith  SRG

     SY is state average sediment yield

Phosphorus loss on a particular SRG for a given application rate would differ
from the state average  in relation  to the sediment yield.

Pesticide Loss

     Use of chemical pesticides (insecticides, fungicides, and herbicides)
in agriculture results  in the  possibility of harmful  impact on environmental
quality (U.S. Department of Agriculture, Agricultural  Research Service,
Control of Water Pollution from Cropland, 1976).    Problems with pesticides
occur when they move away from the site of application but continue to affect
the environment.  Pesticides move from  treated fields  primarily in runoff
water or on sediment carried along in the water.    In  this study areas of
potential pollution problems resulting  from  the movement of pesticides off
treated cropland are explored.

     Pesticides differ widely  in  their  chemical properties, toxicity, and
mode of transport.  These characteristics have been identified in previous
studies and are summarized in  Leonard,  Bailey, and Swank.   Properties affect-
ing pesticide behavior  in soils include solubility, extent of adsorption,
and vapor pressure.  The solubility of  a pesticide determines how rapidly the
compound moves in water.  Adsorption, which is the extent of  attraction
between soil and pesticides in this case, controls the quantity of a pesticide
in soil solution and soil vapor.  Hence, adsorption determines such factors
as pesticide mobility, mode of transport, degradation, and volatility.  For
example, movement of pesticides that  are adsorbed  on  soil colloids is highly


                                       23

-------
 dependent  on  the  rate  of  soil  loss  or  erosion.  Pesticide vapor pressure
 determines potential volatility from s.oils.

      In  identifying potential  pesticide pollution problems,  toxicity  and bio-
 magnification must be  taken  into  consideration.  Because this  study focuses
 on water quality, the  only form of  toxicity  to be considered is toxicity in
 the  aquatic environment.  Hence,  toxicity is measured by the amount of  pesti-
 cides required  to kill fish  in a  specified period of time.   Some pesticides
 can  accumulate  or be magnified in the  biological food chain.   Biomagnifica-
 tion occurs when  some  organism accumulates a pesticide in body tissues  many
 times that in the surrounding  environment (Kenaga, 1975).

      The length of time that a pesticide can persist in the  soil and  aquatic
 environment varies widely from one  pesticide to the next.  Persistence, how-
 ever, depends on  degradation,  adsorption, movement and transport of the
 chemical pesticide.  Pesticide degradation or decomposition  can be accom-
 plished  by microbial as well as nonbiological processes (Armstrong and  Konrad,
 1974; Crosby, 1970; and Kaufman,  1974).

      Factors  concerning application of pesticides must also  be taken  into
 consideration in  identifying potential pollution problems.   The quantity of
 pesticides moving into water obviously depends on the amount of pesticides
 initially  applied.  Both  the application rate per acre and the number of
 acres treated affect the  quantity of pesticide loss.  Because  the quantity of
 a  pesticide moving into water  depends  on the intensity and duration of  rain-
 fall, the  timing  of application is  important.  Those pesticides applied dur-
 ing  rainy  seasons would have a greater potential for pollution  problems.
 Pesticide  losses  are generally highest in the first runoff occurring after
 application of  the pesticide,  and the magnitude of the loss  decreases as time
 between  application and runoff increases (U.S. Department of Agriculture,
 Agricultural  Research  Service, Control of Water Pollution from Cropland,  1976).
 The mode of application affects pesticide loss.  For example,  soil-incorpo-
 rated pesticides  will  not be lost in runoff to as great an extent as pesti-
 cides sprayed on  foliage.

      The approach used in this study has two components:  (1)  How harmful is
 the pesticide?, and (2) How much  of the pesticide can potentially move  from
 applied  sites to  harm  environmental quality?  Both of these  components  cannot
 be analyzed as  simply  as  other pollutants from agriculture.  In the first
 place, a large  number  of pesticides are utilized in agricultural production.
 Furthermore,  several dimensions of  the first component have  environmental
 relevance - toxicity,  persistence,  and biomagnification.  These dimensions
 combined with the number of pesticides result in a large amount of data on
pesticide use.  If the  results of the analysis are to be useful for policy
makers concerned with  changes  in  land use, these data must be  condensed—yet
in a manner that maintains their  environmental relevance.

     Analysis of the first component is complicated due to the fact that prior
research on the processes of pesticide movement in the environment is lacking.
For the analysis in this study, no  standard loading function analogous  to
functions for sediment, nitrogen, and phosphorus was available for pesticides.
                                      24

-------
Each of  these problems required special attention in development of methodo-
logy.

     Previous researchers have developed indices to study the harmful impact
of various  pesticides in the environment.  Klaus Alt reviewed two such indices
before developing one for his study.   One index was defined as follows.

                               PER =MxLxTxB                         (8)

where PEH = potential environmental hazard, M = mobility of the pesticide,
L = longevity of  life, T = toxicity to fish, and B = biomagnification.   Each
of these factors  are given an ordinal rating to construct an index.   An  alter-
native index that is specific to management of a specific crop rather than  a
particular  pesticide is:
                                1 = S  d  x ID                             (9>
                                    .  d.  x LDj


where  I = value  of  the index,  A.  = amount of pesticide  j  used,  dj = decompo-
sition rate  of the  pesticide j j-'and LD. = lethal dose of  pesticide j  to test
animals.  Alt adapted these indices in^his analysis  to  the  following  form.

                                       h.  x A.

                             EEI = I  in  2 X'LD.                          (10>

where  EEI =  environmental  exposure index,  h.  =  half- life  in years of  pesti-
cide i, A. = periodic application rate of pesticide  i,  In 2 = natural log of
2, and LD. = lethal dose of pesticide i to rats.

     The limited number of crops and management systems considered in this
study  precluded  the necessity  to develop  aggregate indices.  The index chosen
could  be specific to particular  pesticides.   In addition, the use of  toxicity
to fish as used  in  the first index was judged to  be more  relevant for non-
point  source water  pollution concerns.  The application rate factor in the
second and third indices,  however,  appeared to  be a useful  addition;  the
amount of pesticide applied is a relevant  factor  in judging its potential
harm.  With  these considerations the index used in this study is:

                             I.  = A.  x T.  x P.                            (11)
                              J     J    J    J

where  A = quantity  applied per hectare, T  = toxicity to fish, and P = persis-
tence  in soil.   Each of  these three factors is  given a  qualitative rating,
based  on secondary  data — A and T  have values of 1, 2, 3 and P has values of 1,
2, 3,  or 4.   The pesticides are  then ranked on  the basis  of their Ij  values
into four categories of  potential impact.  (Details of the quality ratings for
the factors  and  overall  impact are  presented in the empirical section.)

     As with all indices,  some information on the potential environmental
harm of pesticides  is  lost in the construction  of indices.  It was judged,

                                       25

-------
however, that some condensation of information was necessary for meaningful
policy decisions.  Without such condensation, decision-makers would be  over-
whelmed with the quantity of technical data.  If more details concerning
a specific pesticide are desired in the decision process, such information
could be developed with the information provided by this methodology.

     Methodology for consideration of the second component of the analysis
of pesticides (i.e., their potential movement) was even less satisfactorily
developed in the literature.  The loading functions for pesticides presented
by McElroy and coworkers were not satisfactory for this study.  These func-
ions consider concentration of pesticide in the soil to be an important
factor.  Calculation of this concentration requires knowledge of both the
initial concentration and the impact of pesticide applications on concentra-
tion.  Limited secondary information exists on both these parameters.  Par-
ticularly limiting was information on initial concentration of pesticides on
land recently converted to cropland.  In fact, McElroy and coworkers recom-
mended development of site specific data on concentration.  Because sample
collection and laboratory analysis were beyond the scope of this study, the
actual loadings of pesticides could not be estimated.   An alternative approach
was to present potential loadings.  This approach was  based on presentation of
estimates of quantities of pesticides applied, summarized with the index pre-
sented earlier in this subsection for different provinces of the state.  In N
addition, qualitative information was considered in reference to mode and time
of application, factors that do influence potential pesticide loadings.

     One of the major factors found to directly influence the total amount of
pesticide lost, and the resultant concentration in runoff and sediment, is
the proximity of runoff-producing rainfall to the application date.   Charac-
teristically, pesticide losses are highest in the first runoff occurring
after application of the chemical, and the magnitude of the loss generally
decreases as time between application and runoff increases.   The effect of
time is particularly noticeable for short-lived pesticides and for pesticides
that are not incorporated into the soil.  Concentrations of chemicals in sub-
sequent runoff events decrease at a rate that depends  largely on the persis-
tence of the pesticide in the soil.   Normal planting date for selected crops,
mode and time of pesticide application, and monthly percipitation and river
flow data can be used along with pesticide characteristics to evaluate the
potential movement of pesticides away from site of application.


ECONOMIC IMPACTS

     The methodologies used to estimate the economic benefits and costs asso-
ciated with increased crop production are described in this section.  Economic
factors to be considered are discussed in Figure 4.  The most obvious bene-
its are increased farm sales and net farm income.  In addition,  the aggregate
economy benefits as a result of the larger volume of farming inputs purchased
from the nonfarm sector and the larger volume of farm products being pro-
cessed and handled by agribusiness firms.   On the other hand, however, social
costs associated with increased crop production occur as increased quantities
of fertilizers,  pesticides,  and sediment are released to the environment.
                                      26

-------
Net Farm Income

     This component of  the methodology  provides measures of both the positive
and negative impacts  on profits  of  changing  agricultural land use.  To deter-
mine the net increase in profits on land  taken out  of forest or pasture to
be placed in crop production,  the opportunity costs associated with foregone
profits from production of timber or pasture products must be taken into con-
sideration.  Therefore, estimates of net  income per hectare for crops,
forest, and pasture are accounted for in  the methodology.

     Enterprise budgets were  developed  for major  crops produced on new crop-
land.  This approach  would exclude  crops  with acreages limited by government
programs and emphasize  feed grains  and  soybeans,  which increased during the
1973-1976 period in response  to  expanded  export demand.

     Budgeted costs included  expenditures for seed, labor, fertilizer, lime,
chemicals, tractors,  machinery (fuel, lubricants, and repairs), interest, and
custom operations.  Costs were based on the  average use of inputs and average
annual prices.  For example,  the fertilizer  use per hectare was multiplied
by its price to determine the fertilizer  cost.  Estimated variable costs for
tractor and machinery items were based  on average use to till and harvest one
hectare of the specified crop.   Interest  on  operating capital was charged at
nine percent for the  length of the  production period.  The operating capital
was assumed to be borrowed only  for the length of time necessary to secure
inputs in a timely fashion.   The actual loan for  the money was assumed to be
closed before the production  period began.   Several factors, including com-
bining, hauling and storage,  are related  to  crop  yields.

     Crop yields in each county  were calculated for every SRG producing crops.
Two sources of information were  used:   (1) potential yields for each SRG as
reported by the Soil  Conservation Service (U,S. Department of Agriculture,
Soil Conservation Service, Compilation  of Soils Data in Georgia, 1972); and
(2) county average yield as reported by the  Economics, Statistics, and Cooper-
atives Service (U. S. Department of Agriculture,  Statistical Reporting Ser-
vice, Agricultural Prices, 1977),

     These budgeting  procedures  embodied  several  assumptions that seemed
reasonable for the situation  in  this study.  First, the use of average
annual prices for outputs and inputs is appropriate only if the changes in
these quantities were small enough  that they would not affect price.  Given
that the acreages of  concern  in  this study were small relative to existing
state acreages, which are also small relative to  national totals, this
assumption seemed reasonable.  In situations in which the acreage of concern
was larger relative to  aggregate levels,  changes  in consumer and producer
surpluses would need  to be calculated rather than changes in net farm income
(Eckstein, 1958; Currie, Murphy, and Schmitz, 1971).  These surpluses measure
the value of increased  agricultural production to consumers and producers,
respectively.

     The second assumption is that  fixed  costs need not be considered in the
analysis.  Given that small acreages are  involved,  it was assumed that exist-
ing machinery and management  were sufficient for  the incremental cropland.

                                      27

-------
With the well known overinvestment in farm machinery, this assumption is rea-
sonable if the new cropland is being added to existing farm firms.  If  the
new cropland served as the basis for new farm firms, investment in machinery
would have been necessary and the fixed costs of machinery would have to have
been considered in the analysis.  Primary data from farm surveys could  be used
to validate this assumption.

     Procedures for estimating net revenues from forestry and pasture could
not be based on existing enterprise budgets.  In the case of forestry,  the
existing state of knowledge on farm forestry enterprises in the area of study
is nearly void.  This statement refers to both existing enterprise budgets
and secondary data on yields and prices.  Differences in tree species through-
out the area of study combined with differences in management precluded any
systematic development of budgets for this study.

     The source of data utilized to estimate net returns from forestry was
county estimates of gross farm income from forestry sales in the 1974 Census
of Agriculture (U.S. Department of Commerce, 1977).  From these data, it was
possible to derive gross income per hectare of farm forestland in each county.
Because specialized equipment is used for most forestry harvest, it was rea-
sonable to assume that this income reflected only stumpage prices; thus, no
harvesting costs would have to be deducted from the gross sales.  Furthermore,
the establishment and maintenance costs of farm forests can be assumed to be
zero because of the low maintenance costs.  Under these assumptions, the gross
sales could be considered net income from forestry for 1974.   Under the fur-
ther assumption that 1974 was representative of sales levels in the period
of analysis, these sales data could be considered representative of net reve-
nue from forestry.  These methods probably overstated the net revenues from
farm forestry.

     The problem with estimation of net revenues from pasture was different
from forestry in that past agricultural economics research has emphasized pas-
ture budgets (Wise, 1974).  The pasture enterprise in this research, however,
is an intermediate product in beef production.  Adopting past research would
have necessitated also considering beef enterprises.  The problem with beef
enterprise budgeting would have been determination of the representative beef
enterprises.  A large number of different beef enterprises - cow-calf, stocker
cattle, backgrounding calves, and fattening cattle - combined with a large
variance in use of pastures along with other feed in beef enterprises makes
this a difficult process without considerable primary data.   Furthermore,
historical studies of the structure of beef cattle enterprises have doubtful
validity in this study period because of the adjustment of beef  producers to the
low cattle prices during 'this period.  The approach for estimation of net
revenues from pasture therefore was based on secondary data on cash rents for
pasture that is available for Crop Reporting Districts (USDA, ESCS, "Cash
Rents," 1976).  Even if the pasture was not rented, these cash rents do define
opportunity costs for pasture use.  In this study, it was assumed that all
variable costs associated with pasture use such as fence repairs, fertilizer,
and weed control were assumed by the renter.  Prorated fixed costs for fencing,
however,  was deducted from the cash rents to determine net revenue from pas-
ture.   The fencing costs were estimated for a 16.2 hectare rectangular pasture.
                                       28

-------
As in the methodology  for  estimating net income from other land  uses?  the
determination of validity  of these assumptions would require primary data.

Impact on the State's  Economy

     Shifts in  agricultural production and farm income affect not  only farmers
but also income and employment in agribusiness firms that  process  and  handle
agricultural products  and  supply inputs to farming.   The effects of these
changes, however,  do not stop here, as agribusiness  firms  purchase from other
industries and  workers in  agriculture and agribusiness spend their money for
goods and services produced by other industries.

     Analytical procedures to quantify the interindustry flows between agri-
culture and nonfarm sectors are complex and expensive to use.  Input-outout
analysis, for example, requires a great deal of primary data that  must t i
generated through  surveys.  Also, current techniques are focused on the many
sectors of the  nonfarm sector.  This level of disaggregation in  the nonfarm
sector is not needed to measure the impact of agriculture on the  state's
economy.  A major  limitation of techniques currently used  to measure inter-
industry flows  is  that they measure average effects  rather than  marginal
effects.  Thus  an  alternative approach is needed to  measure the  marginal
effect of changes  in agricultural production, which  is more useful than
average effects for guiding investment decisions.

     Adequate secondary data are available for state and sub-state economies
to quantify the effect of  a change in agricultural production on important
economic variables.  An econometric model of interindustry economic flows was
developed for state economies.  Attention was focused on the role  of the
agricultural sector in the overall economy.  Realizing that modeling economic
systems is enormously complex, primary concern was devoted to identifying and
quantifying the principal  economic characteristics of the  economy.  Methodo-
logical development and estimation of this model is  reported in  Appendix B.

     The impact on the state's economy of the changes in agricultural  output
identified in this report  was assessed by using the  interindustry  model pre-
sented in Appendix B.   Land changes in production have resulted  in increased
input usage, agricultural  production, and net farm income.  Each of these
factors was estimated and  entered into a simulator based on the  economic
interindustry model.  The  simulation model used the  estimated farming  vari-
ables by linking  the agricultural and nonagricultural sectors through  a set
of economic flows.  The impact of changes that have  occurred in  agricultural
production was  traced through the economic system by observing changes in
these flows.

Environmental Costs

     This section  addresses the potential social costs that may  result from
converting fores±  and pasture land to crop production.  In recognition of
these social costs, this section discusses the impact of nonpoint  source
pollutant loads on water quality and resultant impacts on  water  uses.   The
                                       29

-------
section also provides a general discussion of the nature of these environ-
mental costs and of the problems of measuring and assigning a value to such
costs.

     Pesticides, fertilizers, and sediment are the primary substances result-
ing from increased crop production that have an impact on water quality.
Although the quantity of sediment is not appreciably reduced in the receiving
water, pesticides and fertilizers are reduced by the biological, physical,
and chemical processes that occur in water.  Sediment can cause turbidity in
surface waters, which makes the water less attractive and can damage fish life
(Kneese and Bower, 1968).  Turbidity may also inhibit the growth of algae and
other aquatic plants.  Increased fertilizer loadings may aggrevate the pollu-
tion problems for water bodies that are susceptible to eutrophication (nutri-
ent buildup).  Thus, nutrient and sediment loss from agriculture can have
visible effects on water quality.  These take the form of algae blooms, murky
or shallow water, and dense weed growth (Ogg, Christensen, and Heimlich,
1979).  Pesticides may lead to destruction of wildlife and may involve health
hazards for humans.  Furthermore, nonpoint source pollutants can affect the
aesthetic enjoyment of environmental and recreational amenities.

     Environmental impacts, in the context of this study, are externalities
arising from agricultural production.  Specifically, agricultural production
causes pollution that directly affects the production and/or consumption pro-
cess of other sectors and the welfare of consumers without intervening mar-
ket transactions (Buchanan and Stubblebine, 1962; Baumol and Dates, 1979).  A
useful classification system for externalities was provided by Baumol and
Dates in their distinction between depletable and nondepletable externalities.
Depletable externalities have individual effects - a specific individual bear-
ing an environmental effect will preclude another individual bearing the cost.
In contrast a nondepletable externality is collectively borne by a number of
individuals.  The importance of this distinction is that depletable externali-
ties are much easier to measure.  In the first place, the individuals affected
can more readily be identified.  Furthermore, the costs of these externalities
can often be measured because the affected individuals have incentives to
undergo expenditures to ameliorate these damages.

     The limited past research on environmental costs therefore are focused on
depletable externalities.  Sediment damages affect water resource functions of
government for which numerous cost estimates have been synthesized.  Damages
considered have included (1) shortened economic life of reservoirs, (2) sedi-
ment damages associated with flooding, C3) costs of clearing and maintaining
drainage ditches, and (4) costs of removing sediment from water for municipal
and industrial uses (Guntermann, Lee, and Swanson, 1975).

     Another component of depletable externalities that has had some empirical
research concerns costs of pesticide damages.  The primary means of measuring
these damages has been to estimate medical costs and workman compensation pay-
ments to humans and veterinarian costs for domestic animals related to pesti-
cides (Langham,Headley and Edwards, 1972).  The medical costs reflected acute
effects of pesticides rather than chronic health effects that were not suffi-
cient to warrant health expenditures or were not directly relatable to pesti-
cide damages.

                                      30

-------
     Another depletable externality  that  has  received  some  attention  concerns
damages to wildlife recreation.   To  analyze these  damages,  studies utilized
the standard recreation demand methodology of inferring benefits from indivi-
dual market expenditures on  recreation.   The  adaptation for these studies is
that pollution affects the quality of  recreation such  that  recreational ex-
periences have less value to participants (Stevens,  1966; Shulstad and
Stoevener, 1978).

     Although a comprehensive study  of the costs of  nonpoint pollution asso-
ciated with new crop production would  have to estimate such costs, methodolo-
gical problems would exist in such estimation.   All  the pollutants from
agriculture of concern in this study  probably  influence wildlife habitats for
at least some species.  Estimating the impact of agricultural pollutants
separate from pollutants from other  sources and  other  disturbances to wild-
life and then relating this  to behavior of recreators, however, would require
a comprehensive multidisciplinary study.

     The problems  of estimating nondepletable environmental costs are even
more difficult.  The contribution of pollution to  deterioration in aesthetics
and damage of ecological systems  typically affect  a  large number of people.
The aspect of biomagnification of pesticides  in  particular  would indicate
possible damage over a long  period of  time.   The collective nature of these
costs makes inference  of their magnitude  from individual market behavior dif-
ficult.  Some economists have approached  this problem  by eliciting preferences
through the use of surveys  (Cicchetti  and Smith, 1976; MeConnell, 1977).  This
approach has questionable validity,  however,  because of its separation from
actual behavior.   In part, the number  of  potential collective benefits from
environmental costs would require a  large number of  separate estimates to as-
certain whether all costs are being  considered.  For example, Krutilla and
Fisher have argued that collective benefits from wildlife exist for several
categories of people other than the  actual participants—those who value pre-
servation of aspects of the  environment,  those who value the option of par<-
ticipating in the  future, and those  who may benefit  from scientific advances
associated with preservation of wildlife.

     A complete, thorough analysis of  all environmental costs resulting from
placing forest and pasture land in crop production is  impossible because it
is too difficult to estimate the  multitude of impacts  of a  change in water
quality (Meta Systems, Inc., 1979).  Even if  some  of these  costs are measured,
political decisions would be necessary to weigh  the  benefits of new cropland
with measurable and nonmeasurable environmental  costs.  Such problems have led
some economists to recommend that all  costs and  benefits be measured  in the
most scientifically valid manner  and presented to  political decision-makers in
such a form (McKean, 1968; Barkley and Seckler,  1972).  From this viewpoint,
the environmental  impacts estimated  in this study  can  be conceptualized as
part of an economic decision framework.   Although  this approach is modest in
its objectives, it does not  use economic  methodology to estimate costs that
are likely to be misleading  and/or incomplete given  the state of the  art.

     The methodology developed in this report was  applied to Georgia  for the
period 1973-1976.  The following  sections report on  this empirical applica-
tion.

                                       31

-------
                                  SECTION 4

                 LAND RESOURCES AND USE PATTERNS IN GEORGIA
     This section summarizes the land resources of Georgia.  SRGs, which are
the basic units of analysis used in this report, are identified and charac-
terized.  LRAs are characterized according to SRGs and land use patterns.
LAND RESOURCE AREAS

     The Southeastern United States contains five LRAs as identified by the
Soil Conservation Service (Austin, 1965), with the three largest extending
into Georgia.  The South Atlantic and Gulf Slope Cash Crop, Forest, and Live-
stock Region makes up almost 75 percent of the state.  Over 90 percent of
Georgia's cropland is found within this region.  However, all regions of
Georgia, which are subdivided into eight LRAs, will be considered in this
study.

     Georgia has a great number of parent materials as well as a wide range
of climatic and topographic conditions that are largely responsible for the
great number of different soils.  The eight LRAs, better known as provinces,
are:  (1) Sand Mountain (LRA 129), (2) Southern Appalachian Ridges and
Valleys (LRA 128), (3) Blue Ridge Mountain (LRA 130), (4) Southern Piedmont
(LRA 136), (5) Carolina and Georgia Sand Hills (LRA 137), (6) Black Lands
(LRA 135), (7) Southern Coastal Plain (LRA 133), and (8) Atlantic Coast
Flatwoods (LRA 153).  The general map of Georgia with physiographic provinces
is shown in Figure 5.  In this report, the two smallest LRAs have been com-
bined with others and therefore the results will be reported for only six
LRAs.  Sand Mountain was included in Southern Appalachian Ridges and Valleys,
and Black Lands was included in the Southern Coastal Plain.

     Total land area occupied by each of the six LRAs within Georgia, as
reported in the CNI of 1967 (U.S. Department of Agriculture, Soil Conserva-
tion Service, 1970), is shown in Table 1.  The state had 7.0 million hectares
in Classes I-III, 1.9 million hectares of marginal land, 2.0 million hectares
of wetlands, and 2.9 million hectares of submarginal land.  Only half of
Georgia's total land area has been classified as suitable for cultivation.
Every land resource area has substantial acreages of marginal and submarginal
land.

     As indicated earlier, this study uses SRG classifications that are based
on the dominant soil series in each soil association.  The SRGs of Georgia
are characterized in Appendix Table Cl.  Some SRGs have the same general


                                     32

-------
EZ3
nrn
Sand Mountain (LRA 129)
Southern Appalachian Ridges and Valleys  (LRA  128)
Blue Ridge (LRA 130)
Southern Piedmont (LRA 136)
Carolina and Georgia Sand Hills (LRA 13?)
Black Lands (LRA 135)
Southern Coastal Plain (LRA 133)
Atlantic Coast Flatwoods (LRA 153)
           Figure 5.  Land resource areas of Georgia.
                            33

-------
                      TABLE 1.   TOTAL CULTIVATABLE,  MARGINAL, WETLANDS AND SUBMARGINAL LANDS
                                           IN GEORGIA BY LAND RESOURCE AREA*
u>
Inventoried Land Area*
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total

Distribution
Cultivatable
(I, II, III)


264,090
83,390
1,848,727
318,838

3,596,706

897,139
7,008,890

50.8
Marginal
(IV)


106,829
62,025
1,019,883
130,618

471,143

103,305
1,893,803

13.7
Wetlands Submarginal
(V) (VI, VII, VIII)
(Hectares)

0
234
15,364 1
46,703

928,198

1,004,182
1,994,681 2
(Percent)
14.4


313,491
315,134
,149,728
149,135

604,704

376,633
,908,825

21.1
Total
Inventory//


684,410
460,783
4,033,702
645,294

5,600,751

2,381,259
13,806,199

92
Percentage
Total of Land Area
Land Area** Inventoried


752,791
652,370
4,394,230
766,488

5,779,257

2,707,043
15,052,179

100
(Percent)

91
71
92
84

97

88
92


    *Source:   U.S.  Department of Agriculture,  Soil Conservation  Service.   Georgia  Conservation Needs  Inven-
     tory,  1967.  Washington, D.C.,  April,  1970.

    //Certain  Federal land,  urban build-up and  water areas were excluded from the inventory.

    **Source:  U.S.  Department of Commerce,  Bureau of the Census.   1969 Census  of  Agriculture.  Washington,
     D.C., 1972.

-------
name, but are different because  of  their slopes.   Capability class  is  based
on dominant soil series and  takes into  consideration average slope.  Other
SRG characteristics reported in  this  table include hydrologic soil  group,
suitability, and erodibility factor.  Appendix Table C2  gives the breakdown
of land area by SRG,  showing the number of hectares of each SRG within the
six LRAs.

     The SRGs have been grouped  by  capability class and  their areas  aggre-
gated for each LRA in Table  2.   This  approach provides a second approxima-
tion of the distribution  of  land area by capability class for the various
LRAs.  The data in Tables  1  and  2 were  compared to determine whether the two
samples were drawn from similar  populations.   Using a variety of non-para-
metric tests for two  related samples  (Siegel, 1956),  the null hypothesis that
the data from the CNI of  1967 (U.S. Department of  Agriculture, Soil  Conserva-
tion Service, 1970) differed significantly at the  10 percent level from the
data developed in this study using  SRGs was not rejected.   While there are
differences for some  individual  categories, the approach used in this  study
appears to be generally consistent  with the approach reported in CNI of 1967
(U.S. Department of Agriculture, Soil Conservation Service,  1970).  Further-
more, the use of the  dominant soil  series to  characterize a soil association
did not introduce an  overall bias in  the data set.
LAND USE PATTERNS

     A thorough inventory  of  land  cover  and land-use  patterns for Georgia is
presented in the CNI  of  1967  (U.S.  Department  of  Agriculture, Soil Conserva-
tion Service,  1970).   A  summary of  the inventory  by LRA is presented in
Appendix Tables C3, C4,  and C5.  Over  13 million  hectares were identified
in row crops,  forest,  conservation  use only, pasture  and rangeland, orchards,
close grown crops, and idle land.   Almost  70 percent  is  in forest and almost
11 percent in  row  crops.

     Crop land is  predominantly on  land  classified as suitable for cultiva-
tion (Appendix Table  C3).  Only 6.3 percent of total  row crop area was on
marginal and submarginal land (includes  wetlands).  However, three LRAs -
Mountains, Piedmont,  and Sand Hills -  had  over 10 percent of their crops on
marginal and submarginal land.   The Southern Coastal  Plain, the major agri-
cultural area  of the  state, had one million hectares  on cultivatable land.
This LRA also  had  the largest area  of  marginal and submarginal land in crop
production, 50 thousand  hectares.   The Piedmont with  21 thousand hectares
has the second largest area of  marginal  and submarginal cropland, followed
by the Atlantic Coast Flatwoods with 12  thousand  hectares.

     The 9.6 million  hectares of forests are more evenly balanced among the
classes (Appendix  Table  C4).   Of Georgia's total  forest acreage, 37.9 per-
cent is on cultivatable  land, 15.4  percent on  marginal land, 20.1 percent
on wetlands, and 26.6 percent on submarginal land.  The Piedmont and Southern
Coastal Plain  each have  over  a  million hectares of forests on land classified
as suitable for cultivation.   In fact, there are  over 3.6 million hectares
of forests on  cultivatable land.


                                       35

-------
                TABLE 2 .   CULTIVATABLE,  MARGINAL,  WETLANDS AND SUBMARGINAL LANDS IN GEORGIA BASED ON
                                DOMINANT CAPABILITY CLASSES ASSIGNED TO SOIL RESOURCE GROUPS
u>
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total

Distribution
Cultivatable
(1,11,111)

157,330
61,696
2,256,335
115,430
4,350,337
1,189,334
8,130,462

54.0
Marginal
(IV)

32,180
0
1,286,755
423,913
801,896
821,548
3,366,292

22.4
Wetlands
(V)
(Hectares)
58,459
20,249
528,667
31,297
94,237
272,086
1,004,995
(Percent)
6.7
Submarginal
(VI, VII, VIII)

504,826
570,425
322,473
195,848
532,787
424,075
2,550,434

16.9
Total

752,795
652,370
4,394,230
766,488
5,779,257
2,707,043
15,052,183

100.0

-------
     Land use patterns of pasture  and rangeland,  conservation use  only,
orchards, close grown crops,  and idle land have all been grouped together
under the heading of Other Uses.   Almost  75 percent of  this  land is  classi-
fied as suitable for cultivation  (Appendix Table  C5).   Only  2 percent  of this
land is classified as wetlands, and  the remaining 25 percent is evenly split
between marginal and submarginal land.  According to this classification
scheme, almost 960 thousand hectares are  classified as  cultivatable  land in
the Southern Coastal Plain, and over 650  thousand hectares classified  as
cultivatable in the Piedmont.  The Piedmont has almost  190 thousand  hectares
classified as marginal land.

     It is expected that a large proportion of  new cropland  will be  converted
from this Other Uses category.  First,  a  large  proportion of this  acreage is
on land suitable for cultivation.  Secondly,  it is more difficult  and  expen-
sive to convert forestland to crop production.  Also, if root  harvesting
equipment is not readily available the  land may have to lie  idle for several
years to allow the tree roots to decay  before the land  can be  put  in cultiva-
tion.
SOIL BANK LAM)

     Total amount of land placed  in  the  Soil  Bank or CRP reached a peak in
Georgia in 1960 with 430,001 hectares  (U.S. Department of Agriculture, Agri-
cultural Stabilization  and  Conservation  Service, Final Report Conservation
Reserve Program:  Summary of Accomplishments  1956-1970, 1970).  Soil Bank
land remained over 430  thousand hectares through 1963, then declined to 320
thousand hectares in 1967,  215 thousand  hectares in 1968, and 100 thousand
hectares in 1969.  Essentially all contracts  had expired by 1970.

     The distribution of Soil Bank land  by LRA in the late 1950 ls is pre-
sented in Table 3.  Over 30 percent  of the Soil Bank land was located in the
Southern Coastal Plain, and 28 percent in the Piedmont.  Twenty-five percent
of Georgia's designated Soil Bank land was in grasses and legumes and 75 per-
cent was planted in trees.  These proportions varied substantially by LRA.
In the Appalachian Valley and Ridges and Mountains more Soil Bank land was
planted to grasses and  legumes than  trees.  Almost 40 percent of the Pied-
mont's designated Soil  Bank land  was in  grasses and legumes.  Only 20 percent
of the Southern Coastal Plain's designated Soil Bank land was in grasses and
legumes.  It would be expected that  areas with a higher proportion of grasses
and legumes might have  relatively more conversion of Soil Bank land to crop-
land, other things being equal.

     The distribution of land restricted to conservation use only by LRA and
soil class is reported  in Table 4.   Total land area in this category for each
LRA is very similar to  the  Soil Bank hectares shown in Table 3.  Over 85 per-
cent of this land is classified as suitable for cultivation, and ten percent
is classified as marginal (Table  4).
                                       37

-------
               TABLE 3.  CONSERVATION RESERVE LAND IN GEORGIA IN 1956-1959 BY LAND RESOURCE AREA*
OJ
oo
Land Resource
Area
Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total
Distribution
Proposed
Trees
1,985
416
50,709
17,267
148,773
7,572
226,722
71.4
Proposed Grasses
and Legumes
(Hectares)
3,092
465
31,162
3?632
38,381
458
77,190
(Percent)
24.3
Other
634
111
6,665
470
5,688
42
13,610
4.3
Total
5,711
992
88,536
21,369
192,842
8,072
317,522
100
     *Source:   U.S.  Department of Agriculture, Agricultural Stabilization and Conservation Service.  Georgia
               ASCS  Annual Reports.  Athens, Georgia, Annual.

-------
 TABLE 4 .   CULTIVATABLE, MARGINAL, WETLANDS AND SUBMARGINAL LAND UNDER CONSERVATION USE ONLY IN 1967*
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total

Distribution
Cultivatable
(I, II, III)

17,469
6,475
98,328
17,534

273,911

30,728
444,445

85.2
Marginal
(IV)

4,712
1,616
24,396
3,048

19,156

1,487
54,415

10.4
Wetlands
(V)
(Hectares)
0
0
61
63

1,482

188
1,794
(Percent)
0.4
Submarginal
(VI, VII, VIII)

2,627
389
9,566
466

7,831

0
20,879

4.0
Total

24,808
8,480
132,351
21,111

302,380

32,403
521,533

100
*Source:  U.S. Department of Agriculture,  Soil  Conservation Service.  Georgia Conservation Needs Inven-
 tory, 1967.  Washington, D.C.,  April,  1970.

-------
                                   SECTION 5

        ESTIMATION OF CHANGES IN LAND USE AND INPUT USAGE IN GEORGIA
     The magnitude and type of land use changes are identified in this sec-
tion.  The type of land use changes considered relate to capability class,
former agricultural use, and the type of crop grown on the land once it is
converted to crop production.  Special consideration is given to conversion
of Soil Bank land to crop production.  Having identified which crops are pro-
duced on the new crop land, it will be possible to measure changes in input
usage, with particular attention focused on fertilizers and pesticides, which
are potentially hazardous to the environment.
CHANGES  IN CROP ACREAGE

     The five principal crops in Georgia are corn, soybeans, peanuts, cotton,
and tobacco.  The combined acreage of these five crops accounts for over 95
percent  of total row crop acreage in the State (U.S. Department of Commerce,
Bureau of the Census, Census of Agriculture, 1974).  Hence, this discussion
will concentrate on these principal crops.  The Southern Coastal Plains
accounts for over 75 percent of Georgia's crop acreage.  Of the other five
LRAs, only the Atlantic Coast Flatwoods accounted for as much as 10 percent
of Georgia's total crop acreage.  The Appalachian Valley and Ridges, Mountain,
and Sand Hills combined have only about 5 percent of Georgia's crop acreage.
Corn accounted for over half of the total crop acreage over the 1973-1976
period;  soybeans accounted for 20 percent of total crop acreage.  Changes in
crop acreage, however, are of particular interest in this study.

     Changes in crop acreages over the 1973-1976 period for each LRA and the
major crops are reported in Table 5.  Total acreage for the five crops in-
creased by 273 thousand hectares.  Corn and soybeans were responsible for this
increase:  corn increased 238 thousand hectares and soybeans increased 98
thousand hectares.   Peanuts and tobacco, which are generally more profitable
than other row crops, are restricted by government acreage programs.  Cotton
acreage declined 73 thousand hectares over the period.  Hence it may be con-
cluded that new cropland was brought into production for corn and soybeans.
SHIFTS IN LAND USE PATTERNS

     The types of land use patterns that had been converted to cropland over
the 1973-1976 period were identified from surveys by the Soil Conservation
Service and the Georgia Cooperative Extension Service.  Their estimates of


                                      40

-------
the percentage distribution of land use patterns were linked with the esti-
mated cropland changes  from this  study to determine the number of hectares
of woodland, pasture  land,  and other land that was placed in crop production
(Table 6).  Most  of the new cropland was converted from woodland.   Approxi-
mately 56 percent of  the land had previously been woodland.   Only 38  percent
of the land placed in crop  production was previously in pasture (Table 6),

     Changes, in land  use patterns on land previously under the Soil Bank
program were identified through the use of two additional surveys - one con-
ducted by the Georgia Forestry Commission and one conducted  as a part of  the
present study.  The Georgia Forestry Commission study was concerned with  the
280 thousand hectares of pine plantations that were established on Soil Bank
land in Georgia between 1956 and  I960 (Georgia Forestry Commission, 1976).
The objective of  that study was to determine the amount of this acreage that
remained in trees.  With the last plantings in 1960, the final payment under
the 10-year contracts was made in 1970.   The average age of  these plantations
was 18 years in 1976.  A random sample of 47 cases was included in the study
for Georgia.  Field foresters examined each tract to determine such factors
as existing growing stock volume, volume removed in the past,  and hectares
converted to other uses.


 TABLE 5.  CHANGES IN LAND  AREA OF FIVE  MAJOR ROW CROPS  IN GEORGIA, 1973-1976*
                    	Row  Crop	
Land Resource Area     Corn   Soybeans    Peanuts    Cotton   Tobacco    Total
(Hectares)
Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
1,859
109
10,203
2,095
9,251
1,150
27,707
3,620
0
0
0
221
-5,530
-644
-17,961
-5,010
0
0
-156
-123
5,580
615
19,793
803
 Southern Coastal
   Plain             202,047   52,607     5,278    -42,864   2,852    219,920

 Atlantic Coast
   Flatwoods          21,218    3,957       184       -519   1,863     26,703

   Total             237,531   98,292     5,683    -72,528   4,436    273,414
 *Source:   U.S.  Department of Agriculture, Statistical Reporting Service,
  Georgia  Crop Reporting Service.  Georgia Agricultural Facts, 1969-1977,
  Athens,  Georgia, October 1978.
                                       41

-------
          TABLE 6.  ESTIMATED CONVERSION OF WOODLAND, PASTURE AND
                    OTHER LAND TO CROPLAND, GEORGIA 1973-1976
                                        All New Cropland
                                               Pasture          Other
Land Resource Area            Woodland           Land           Land
                                               (Hectares)

Appalachian Valley
  and Ridges                     872             3,879           257

Mountain                          14               197             0

Piedmont                       3,741            12,656         4,173

Sand Hills                     2,473             2,531           680

Southern Coastal
  Plain                      126,500            81,831        12,231

Atlantic Coast
  Flatwoods                   29,607            10,519         1,040

  Total                      163,207           111,613        18,381
     Results of the survey indicate that 83 percent of the Soil Bank  land
 originally planted in pines remained in trees through 1976 (Table  7).  The
 average growing stock was almost 64 cords per hectare (Georgia Forestry
 Commission, p. 5, 1976).  Many of the plantations were approaching sawtimber
 size, and some already contained sawtimber.  As these trees approached saw-
 timber size, this land produced some $14 worth of wood per hectare per year
 (Georgia Forestry Commission, 1976).

     The remaining 17 percent of the land in pine plantations had  been cleared
 by 1976 (Table 7).  Only one percent of the land had been placed in crop pro-
 duction by 1976.  Trees on that land had been harvested at an early age and
 the land was converted to crops prior to 1973.  A large portion of the land -
 14.5 percent - had recently been harvested by clear cutting and was idle in
 1976.  Trees on some of this land may have been harvested in order to put the
 land into crop production; however, this conversion process would  not be com-
plete until after 1976.

     A second survey was conducted to determine the amount of Soil Bank land
that was converted from pasture to crops.  Approximately 25 percent of the
Soil Bank land had originally been planted to grasses and legumes.  It would
be expected that more of this land would be converted to cropland  than had

                                     42

-------
     TABLE  7.   STATUS  OF  PINE PLANTATIONS  ESTABLISHED IN GEORGIA UNDER THE 1956-1960 SOIL BANK,
                                    CONSERVATION RESERVE PROGRAM*


Trees planted 1956-1960
Trees remaining 1976
Trees harvested 1976
Recently harvested
Converted back to agriculture
Urban or other development
Status for clearing unknown
Land Area
(Hectares)
280,867
233,120
47,747
40,726
3,089
2,247
1,686
Distribution
(Percent)
100.0
83.0
17.0
14.5
1.1
0.8
0.6
*Source:  Georgia Forestry Commission.   A Report  on  the  Current  Status of the Pine Plantations  Esta-
 blished in the State of Georgia Under  the 1956 to 1960  Soil  Bank, Conservation Reserve Program.
 Macon, Georgia, August 1976.

-------
been the case with the Soil Bank land planted to trees.  First, the conver-
sion process for pastureland is fairly simple and can be completed without a
delay in production.  Secondly, beef prices were considered to be depressed
between 1974 and 1976.

     The study was designed to survey landowners who had participated in the
Soil Bank program.  The list of former participants in the program was stra-
tified by geographic region in the state and a sample was randomly drawn from
each region.  Seventy landowners were surveyed to determine the status of
their Soil Bank land in 1976.

     The results of the survey indicated that 87.5 percent of the land
planted to  trees under the Soil Bank program remained in trees in 1976.  The
remaining acreage that was originally planted to trees (12.5 percent) had been
harvested but remained idle in 1976.  These figures are very similar to the
Georgia Forestry Commission estimates.  The survey also indicated that approx-
imately 70 percent  of the land planted to grasses and legumes under the Soil
Bank program had been converted to crop production by 1976.  Corn was the
dominant crop planted on former Soil Bank land.
LAND USE  CHANGES BY CAPABILITY CLASS

     New  cropland, which is identified as an increase in cropland over the
1973-1976 period, is estimated in this section according to specific SRGs
and thus  to capability classes.  Two alternative approaches were considered.
First, LANDSAT remote sensing data were used.  Secondly, it was assumed that
no change would occur in the distribution of cropland among the suitable,
marginal, and submarginal for cultivation classifications from the CNI of
1967 (U.S. Department of Agriculture, Soil Conservation Service, 1970).  The
following paragraphs discuss the results obtained with these approaches.

     Arrangements were made to have the LANDSAT digital data analyzed using
the specialized computer facilities at the Georgia Institute of Technology,
CLANDSAT data were processed on the Earth Resources Data Analysis System
and the Georgia Tech Cyber 74.)  The Engineering Experiment Station at Georgia
Tech contracted to machine process LANDSAT digital data for the determination
of land cover changes.  The analysis was to result in land cover maps with
overlay capacity.  The major responsibilities of researchers at the University
of Georgia included acquisition of all LANDSAT tapes, ground truth acquisi-^
tion, and determination of training field locations and geographic control
points on the computer system.  Initially, change detection was planned for
the period 1972 to 1976, but because no good cloud-free imagery was available
in 1972 (only one satellite was in orbit at that time), computer compatible
tapes covering four LANDSAT scenes in Georgia were ordered for 1973 and 1976.

     After the appropriate ground truth information had been acquired and
entered into the computer, processing of the tapes was initiated.  However,
several problems as described below were encountered.  Finally, it was deter-
mined that land use changes could not be identified accurately using these
procedures.
                                     44

-------
     The best IANDSAT tapes  that  could  be identified during mid-summer had
some cloud cover that affected  quality  of the imagery.   There  is generally
less cloud cover in Georgia  during winter.   Also, with mid-summer scenes
crops were at so many different stages  of growth that it was almost impossible
to classify row crops into one  homogeneous group.   Because both the 1973 and
1976 scenes were taken before this study was begun,  we could not know what
stage of growth the row  crops were in.   Furthermore, LANDSAT technology is
not fully developed.  NASA recommended  that the  LANDSAT  scenes be geo-
referenced before they were  classified.   This process failed and the scenes
had to be classified before  they  were geo-referenced.

     Although several difficulties were encountered in using LANDSAT imagery,
the approach offers a great  deal  of  potential for detecting land use change.
Perhaps the experiences  and  failures of LANDSAT  technology encountered in
this study will be helpful in future efforts.

     With the second approach,  new cropland was  distributed among suitable,
marginal, and submarginal classifications according to the distribution of
land already in cultivation.  Because  the observed  land  use changes for the
1973-1976 period were small  relative to total land  area  in each county, there
was sufficient land area in  each  of  these classifications to meet these
requirements.  The following discussion of land  use changes by capability
classes is based on this approach.

     New cropland estimates  are reported in Table 8. For the  1973-1976
period, 293 thousand hectares of  agricultural land  were  shifted to crop pro-
duction.  Over 12 percent of the  new cropland was on land classified as mar-
ginal or submarginal.  Another  12 percent of the new cropland was on land
classified as suitable for cultivation  and formerly under the  Soil Bank pro-
gram.  Thus 25 percent of all new crop  production (74.5'  thousand hectares)
was on marginal, submarginal, and/or Soil Bank land.

     Data in Table 8 show the distribution of marginal,  submarginal, and
Soil Bank land by LRA.   Eighty-five  percent of the  marginal and submarginal
land brought into crop production was in the Southern Coastal Plain and
Atlantic Coast Flatwoods,  Most of the  Soil Bank land classified as suitable
for cultivation (85 percent), however,  was in the Piedmont and Southern
Coastal Plain.  Much of  the  Atlantic Coast; Flatwoods that was placed in crop
production during this period was wetlands,  which are classified as submar-
ginal land in their natural  state.   With proper  drainage, however, these soils
can be quite productive.

     Over half of the marginal, submarginal and  Soil Bank land placed in
crop production (41.1 thousand  hectares) was in  the Southern Coastal Plain.
In the'Piedmont and Atlantic Coast Flatwoods, 12 thousand and  14 thousand
hectares, respectively,  were placed  in  crop production on marginal, submar-
final, and Soil Bank land.   The three remaining  LRAs each had  less than 4,000
hectares of marginal, submarginal, and  Soil Bank land placed in crop produc-
tion.
                                     45

-------
            TABLE 8.  ESTIMATED LAND AREA CONVERTED TO CROP PRODUCTION BY LAND RESOURCE AREA,
                                           GEORGIA 1973-1976
Land Resource Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total


4,490
176
17,496
3,593
201,847
28,039
255,641

518
35
3,073
2,089
18,715
13,128
37,558
Soil Bank
(Hectares)

3,265
103
9,011
1,126
22,405
1,057
36,967

303
0
993
216
1,826
487
3,825
Non-Soil Bank


1,225
73
8,485
2,467
179,442
26,982
218,674

215
35
2,080
1,873
16,889
12,641
33,733
Total State
293,199
40,792
252,407

-------
CHANGES IN INPUT USAGE

     Considering that cotton acreage declined and peanut and  tobacco  acreages
were restricted by government programs,  it was assumed that land was  brought
into crop production to  produce corn and/or soybeans,  which experienced  tre-
mendous increases in acreage between 1973 and 1976.  Therefore, input usage
was estimated for corn and soybeans.

     The enterprise budgets for crops produced on new  cropland developed in
this study used updated  estimates of previously published production  costs
(McArthur, 1971; U.S. Department of Agriculture, Economic Research  Service,
Commodity Budgets, 1975;  U.S. Congress,  1977; Paxton,  1977; and Wise, 1974).
These budgets, were reviewed for accuracy by personnel  in the  Economics,  Sta-
tistics, and Cooperatives Service, U.S,  Department of  Agriculture and Coopera-
tive Extension  Service.   Representative budgets for corn and  soybeans are
shown in Tables  9  and  10.  These budgets reflect state average yields; how-
ever, in subsequent  analyses yields are estimated for  each SRG.  Production
costs were also  assumed  to vary with yields.  Land with below-average yields
was assumed  to  have  production costs that were slightly below the average.
Each 10 percent  reduction in yields below the state average was assumed  to
result  in a  7 percent  reduction in variable costs.  This assumption may  affect
net farm income  estimates on marginal and submarginal  cropland, but it would
have no effect  on environmental loadings.

     State specialists from the Cooperative Extension  Service in Georgia
helped  estimate  the  amounts of herbicidesf insecticides, and  fungicides used
on corn and  soybeans.   The procecure used to estimate  application rates  for
these chemicals  involved updating the Pesticide and General Farm Survey  that
was conducted  in 1971  (USDA, Economic Research Service, 1971).  Lists of
chemicals identified in that survey along with amounts used per acre  in  1971
were sent to state specialists to estimate the type and amount of chemicals
used in 1976.   This  approach was used to determine representative usage
rather  than  recommended usage.,  A comparison of the 1971 survey results with
estimated usage in 1976 reveals that pesticide usage has generally  increased
since 1971.   Also,  a much greater variety of chemicals was being used in
1976,   Names of  pesticides and quantities used in Georgia will be presented
in the  next  section.
                                      47

-------
     TABLE 9.   REPRESENTATIVE INPUT USE AND COSTS  PER HECTARE OF CORN
                             FOR GEORGIA, 1976
Item
Labor
Seed
Fertilizer, 5-10-15
Anhydrous ammonia
Lime
Herbicide
Tractor
Machinery
Interest on operating
capital (6 months)
Total preharvest costs
Combining
Hauling
Total harvest costs
Preharvest and harvest
costs
Unit
hr.
kg.
kg.
kg.
MT
kg.
hr.
-
dol.
dol.
ha.
kg.
dol.

dol.
Quantity
11.98
12.57
671.62
112.08
0.74
2.80
9.88
-
217.17
-
1.0
3879.53*
-


Price
2.62
1.34
0.11
0.33
17.05
5.98
2.30
-
0.09
-
39.54
0.001
-


Cost
31.41
16.90
70.72
37.06
12.65
16.80
22.73
8.90
19.55
236.72
39.54
3.06
42.60

279.32
*Average yield per hectare of corn for Georgia,  1976.
                                    48

-------
 TABLE 10.   REPRESENTATIVE INPUT USE AND COST PER HECTARE OF SOYBEANS
                           FOR GEORGIA, 1976
Item
Labor
Seed
Inoculant
Fertilizer, 0-10-20
Lime
Herbicide
Pesticide
Tractor
Machinery
Interest on operating
capital (6 months)
Total preharvest costs
Combining
Hauling
Tractor and equipment
Total harvest costs
Preharvest and harvest
costs
Unit
hr.
kg.
kg.
kg.
MT
kg.
kg.
hr.
-
dol.
dol.
ha.
kg-
hr.
dol.
dol.
Quantity
8.40
60.49
0.28
447.75
0.74
1.23
3.36
5.78
-
162.62
-
1.00
1568.90*
1.85
-
-
Price
2.62
0.29
NA
0.13
17.05
13.53
0.77
1.89
-
0.09
-
37.06
0.003
2.31
—
-
Cost
22.02
17.57
NA
58.32
12.65
16.66
2.59
10.92
21.89
14.64
177.26
37.06
4.62
4.27
45.95
223.21
*Average yield per hectare of soybeans for Georgia, 1976.
                                   49

-------
                                  SECTION 6

        ESTIMATED ENVIRONMENTAL IMPACT OF LAND USE CHANGES IN GEORGIA
     The environmental impacts of placing marginal, submarginal, and Soil Bank
land in crop production are described in detail in this section.  Such factors
as sediment, fertilizer, and pesticide loadings are quantified so that they
can be compared in a trade-off fashion with income and employment resulting
from the increased production.
SEDIMENT AND RUNOFF DATA

     The data used to calculate runoff and sediment yield are discussed in
the following paragraphs.  Estimates of soil and cover conditions are used
with storm rainfall data to calculate potential runoff.  Estimates of cli-
matic, soil, topographic, and management practices are used to determine ero-
sion rates.  Erosion rates are then multiplied by sediment delivery ratios
to determine sediment yield.

     Data sources and procedures for calculating annual potential runoff are
presented in Figure 6.  Runoff curve numbers (Appendix Table Dl), which are
related to soil and cover conditions, are used to calculate potential abstrac-
tion (S) for the major agricultural uses - crops, forest, and pasture.
Potential abstraction is then used to calculate potential direct runoff for
1, 2, 5, 10, 25, 50, and 100 year storms.  Annual potential direct runoff
is then calculated as the weighted average of direct runoff for each of these
storm events, with the weights based on the probability of occurrence for each
storm.

     Sediment loading calculations and data sources are presented in Figure 7.
Storm soil losses from a cultivated field are directly proportional to an
interaction term, energy-intensity (El), which is the product of rainfall
energy and the maximum 30-minute intensity.  The El values are summed to
obtain an annual rainfall-erosivity index for a given location.  Georgia's
annual index, which serves as the R factor, is shown in Appendix Table D2.

     The most significant soil characteristics affecting soil erodibility are
texture, organic matter content, soil structure, and permeability.  The K
values, which have been assigned to different soils by the Soil Conservation
Service, are shown in Appendix Table Cl for the 74 SRGs identified in this
report.
                                      50

-------
  Land Use
  Hydrologic
  Soil-Cover
Characteristics
 from Appendix
   Table Dl
      Potential Abstraction

         q _  i>ooo  _  10
         S •   CN      10
                               Storm Rainfall
                                (Rfi) for ith
                           Storm Event (i = 1, 2,  5,
                           10, 25, 50, and 100 Year
                           Storms) from the National
                                Weather Service
                         Potential  Direct  Runoff
                             from ith Storm
                                     - 0.2S)2
                                   i  + 0.8S
                     Annual  Potential  Direct  Runoff
                                   n
                            AQ  =   I     Q
Figure 6.  Schematic  showing  calculation of  annual potential direct runoff,
                                   51

-------
            Soil
         Properties
Slope Lengths,
  and Slope
  Gradients
Land Use Types,
 Canopy Ground
 Cover Density
  Types of
Conservation
  Practice
 Local Drainage
   Area and
  Soil Texture
       Soil Erodibility
         Factor K,
        from Appendix
          Table Cl
Ul
N3
 Topographic
  Factor LS,
from Appendix
  Figure Dl
 Cover Factor C,
 from Appendix
   Table D3
   Practice
Factor P, from
   Appendix
   Table D4
    Sediment
 Delivery Ratio
Sj from Appendix
    Table D5
                                         SY =  (R  • K • LS • C • P) Sf
                                                     Local
                                               Rainfall Erosivity
                                                  Factor R from
                                                Appendix Table D2
                  Figure 7.  Schematic showing calculation of sediment loading from surface  erosion.

-------
     Slope length  (L). and  slope gradient (S) are combined into a single topo-
graphic factor, LS.   The LS factor for gradients up to 60 percent and slope
lengths to 600 meters is obtained from the Slope-Effect Chart  Appendix
Figure Dl.

     The plant cover  or cropping management factor (C) relates to the effect
of cover on soil loss.   The basic soil loss is the rate at which the  field
would erode if it  were continuously in tilled fallow.   The equation's factor
C indicates the percentage of this potential soil loss that would occur if
the surface was partially  protected by some particular combination of cover
and management.  The  C values used in this study are reported in Appendix
Table D3.

     The erosion-control practice factor (P) accounts  for control practices
that reduce the erosion potential of runoff.  Contouring influences drainage
patterns, runoff concentration, and runoff velocity, whereas terracing only
reduces the length of the  slope.  Control practice values for contouring
and terracing are  shown in Appendix Table D4.

     Estimates of  sediment yield are based on sheet and rill erosion  as cal-
culated by the USLE and the sediment delivery ratio as determined by  texture
of the soil and size  of drainage area.  Sediment yields were calculated for
new cropland on each  SRG,  LRA, and the entire state.

     The smallest  drainage areas that produce a perennial stream within the
various LRAs, as shown in  Appendix Table D5, were used to compute sediment
delivery ratios.   Delivery ratios for the various LRAs are also  shown in
Appendix Table D5.  All SRGs falling within Sand Mountain (LRA 129),  Southern
Appalachian Ridges and Valleys (LRA 128), Blue Ridge Mountain (LRA 130), and
Southern Piedmont  (LRA 136) above the fall-line, were  considered to have a
medium textured topsoil.   Those below the fall-line with Carolina and  Georgia
Sand Hills (LRA 137), Black Lands (LRA 135), Southern  Coastal Plain (LRA 133),
and Atlantic Coast Flatwoods (LRA 153) were considered to have a sandy tex-
tured topsoil.

     Fertilizer losses are assumed to be related to empirical estimates  of
runoff and sediment yield.   Because no state estimates were available  to
determine what percentage  of applied fertilizers was lost by runoff and
erosion, national  estimates were used.  It was believed that this approach
would provide conservative estimates of the quantity of fertilizer losses.
In the present study, it is assumed that, on the average, 6 percent of applied
nitrogen is lost by erosion, 2 percent is lost by runoff, and 2  percent moves
through the soil to groundwater.  No attempt was made  to keep track of ground-
water pollution, however,  as the principal focus of this study was to  account
for pollution in surface water.  Land with an erosion  rate that  differs  from
the state average  was assumed to lose proportionately  more or less nitrogen
than the 6 percent state average linked to erosion. Also, it is assumed that
nitrogen loss on land with a runoff rate that differed from the  state average
would differ proportionately from the 2 percent linked to runoff.
                                       53

-------
     It is assumed that approximately 15 percent of the phosphorus applied as
chemical fertilizer on cropland would be lost on the average.  Since phosphate
sorbs strongly on soil and resists leaching, phosphorus loss is closely re-
lated to erosion.  It is assumed that cropland with an erosion rate equal to
the state average would lose 15 percent of the applied phosphorus, whereas
cropland with higher erosion rates would lose proportionately more phosphorus.
Likewise, land with an erosion rate below the state average would lose pro-
portionately less than 15 percent of its phosphorus applied as chemical fer-
tilizer.

     The following subsections report empirical estimates of environmental
loadings from placing marginal, submarginal, and Soil Bank land in crop pro-
duction during the 1973-1976 period.  Similar estimates for all new cropland
are reported in Appendix Tables El through E5«
RUNOFF AND SEDIMENT YIELD

     Water quality can deteriorate as the result of agricultural pollutants
being transported from fields to water bodies.  The two principal transport
processes, as mentioned earlier, are direct runoff and sediment movement.
First, water quality is affected by the mere presence of soil in suspension.
In addition, as sediment is deposited in streams or lakes, the use and
appearance of these water resources may be adversely affected.  Secondly,
water and sediment may carry pesticides and plant nutrients, which also
deteriorate water quality.

     Changing land use from trees or grasses to crops increases runoff and
erosion.  There is only a small increase in runoff, as evidenced in Table 11.
Placing marginal, submarginal, and Soil Bank land in crop production increases
runoff 1.93 centimeters per hectare on the average.  Rather than concentrat-
ing on the change in runoff, it is probably more important to look at the
actual level of runoff from cropland because runoff affects sediment, ferti-
lizer and pesticide loadings.  Table 11 presents estimated runoff from the
new cropland.  Average runoff for new cropland on marginal, submarginal and
Soil Bank land is 9.35 centimeters per hectare annually.  Runoff is highest
in the Atlantic Coast Flatwoods.

     Placing land in crop production generally results in higher annual ero-
sion and hence higher sedimentation than would occur if the land had remained
in forest or pasture.  Increased sediment yields as a result of the shift in
land use (Tables 12 and 13) indicate that sediment from agriculture may
become a significant problem in the future unless better land use planning
and good management are utilized.  Sediment yield from new cropland has in-
creased over 20 metric tons per hectare in three LRAs - Appalachian Valley
and Ridges, Mountain and Piedmont.  Of the three, the Piedmont had the
largest amount of new cropland, resulting in increased sediment yield of over
300,000 metric tons annually.  The largest quantity of sediment from the new
cropland, however, occurred in the Southern Coastal Plain, which has over
400,000 metric tons annually.  It can be noted that this results from the
large amount of new cropland, as- the average per hectare was only 1Q.Q4 metric


                                      54

-------
                    TABLE 11.  ESTIMATED RUNOFF FROM NEW CROPLAND ON MARGINAL, SUBMARGINAL
                                         AND SOIL BANK LAND IN GEORGIA, 1976
Ui
Marginal and
Submarginal Land*
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
State Total

Runoff
(cm)
9.04
9.60
7.62
5.44

8.61

12.24
9.63
Increased
Runoff
(cm)
0.64
0.89
1.30
2.62

2.51

3.35
2.69
Cultivatable
Soil Bank Land

Runoff
(em)
6.73
9.91
9.43
10.74

9.04

11.38
9.07
Increased
Runoff
(cm)
1.14
1.50
0.36
1.02

1.42

2.41
1.17
Marginal,
Submarginal and
Soil Bank Land

Runoff
(cm)
7.04
9.85
8.97
7.29

8.86

12.17
9.35
Increased
Runoff
(cm)
1.09
1.35
0.58
2.06

1.93

3.28
1.93
        ^Includes marginal and submarginal Soil Bank land..

-------
             TABLE 12.  ESTIMATED INCREASED SEDIMENT YIELD FROM NEW CROPLAND ON MARGINAL,
                           SUBMARGINAL, AND SOIL BANK LAND IN GEORGIA, 1976
Marginal and
Submarginal Land*
Land Resource
Area

Appalachian Valley and
Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
State Total
Average
Total per Hectare
(1,000 MT)
16.16
2.11
96.96
16.93
193.52
8.68
334.36
(MT)
31.20
60.44
31.56
8.11
10.35
0.67
8.92
Cultivatable
Soil Bank Land
Average
Total per Hectare
(1,000 MT)
68.88
2.35
25.34
8.54
218.84
1.74
525.69
CMT)
21.10
22.80
25.02
7.59
9.77
1.64
14.22
Marginal,
Submarginal and
Soil Bank Land
Average
Total per Hectare
(1,000 MT)
85.04
4.46
322.30
25.48
412.36
10.42
860.06
(MT)
22.49
32.41
26.68
7.93
10.04
0.74
11.54
*Includes marginal and Submarginal Soil Bank land.

-------
          TABLE 13.  ESTIMATED  SEDIMENT YIELD FROM NEW CROPLAND  ON MARGINAL,  SUBMARGINAL,
                                AND  SOIL  BANK LAND IN  GEORGIA, 1976
Marginal and
Submarginal Land*
Land Resource
Area

Appalachian Valley and
Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
State Total
Average
Total per Hectare
Cl,000 MT)
16.71
2.17
99.91
17.59
203.29
9.21
348.88
(MT)
32.26
62.07
32,52
8.42
10.86
0.69
9.30
Cultivatable
Soil Bank Land
Average
Total per Hectare
(1,000 MT)
70.84
2,41
231,78
8.79
225,06
1.78
540.66
(MT)
21.71
23.45
25.74
7.82
10.06
1.68
14.67
Marginal,
Submarginal and
Soil Bank Land
Average
Total per Hectare
(1,000 MT)
87.54
4.58
331.69
26.38
428.35
10.98
889.52
(MT)
23.14
33.26
7.46
8.20
10.42
0.78
11.94
*Includes marginal and Submarginal Soil  Bank land.

-------
tons.  Sediment yields per hectare as reported in Table 13 are used in cal-
culating nitrogen and phosphorus loss.
FERTILIZER LOSS

     Estimates of annual nitrogen losses from crop production in 1976 on new
cropland on marginal, submarginal, and Soil Bank land vary considerably by
LRA (Table 14).  Even ignoring the Mountain LRA, which had only minimal land
area of the new cropland, nitrogen losses ranged from 3.98 kilograms per
hectare in the Atlantic Coast Flatwoods to 11.00 kilograms per hectare in
the Piedmont.  The pollution potential from nitrogen fertilizer is greatest
in the Southern Coastal Plain where large areas receive high rates of ferti-
lizer.  This LRA alone accounts for 61 percent of nitrogen losses from the
new cropland.

     Similarly, aggregate phosphorus losses from marginal, submarginal, and
Soil Bank land, as reported in Table 15, are highest in aggregate in the
Southern Coastal Plain (403 thousand kilograms annually) and the Piedmont
(273 thousand kilograms annually).  Phosphorus losses on a per-hectare basis,
as shown in Table 15, are closely linked to sediment yield reported in Table
13.  The new cropland in the Atlantic Coast Flatwoods lost very little phos-
phorus, whereas the other five LRAs had average phosphorus losses over 6
kilograms per hectare.  This quantity of phosphorus loss is higher than the
previously published estimates that were reviewed in an earlier section of
this report.  Most of those studies concerned all farmland, including forest
and pasture land.  Consequently, average erosion would be much less than the
erosion rates from cropland on marginal, submarginal, and Soil Bank land in
Georgia.  Also, phosphorus application rates in Georgia are relatively high
compared to the national average.
PESTICIDE LOSS

     As indicated earlier, most of the increased row cropland in Georgia
during the 1973-1976 period can be attributed to corn and soybeans.  Several
dozen destructive pests of corn and soybeans cause heavy losses virtually
every year in Georgia.  The names of these major pests can be seen in Table
16.  This information was compiled from Georgia Agricultural and Entomology
Extension Service Bulletins.

     The pesticides identified by state specialists in the Georgia Cooperative
Extension Service as being applied to corn and soybeans have been charac-
terized by application rate,  toxicity to fish and persistence in soil in Table
17.  These pesticides are given a rating for each of the three factors.  Pest-
icides are assigned the lowest rating for low application rate, relatively
nontoxic and nonpersistent characteristics.  Detailed guidelines for rating
the various factors appear as footnotes in Table 17.

     An index of potential environmental impact is calculated as the product
of the ratings for quantity used, toxicity, and persistence.  The value of


                                      58

-------
               TABLE 14.   ESTIMATED NITROGEN LOSS FROM NEW CROPLAND ON MARGINAL, SUBMARGINAL, AND SOIL
                                                 BANK LAND IN GEORGIA, 1976
Ul
Marginal and
Submarginal Land*
Land Resource
Area

Appalachian Valley and
Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
State Total
Total
(1,000 Kg)
4.62
1.82
44.81
8.94
156.18
52.14
268.51
Average
per Hectare
(Kg)
8.94
52.01
14.60
4.28
8.36
3.98
7.16
Cultivatable
Soil Bank Land
Total
(1,000 Kg)
23.14
2.29
87.97
5.93
213.50
4.16
336.99
Average
per Hectare
(Kg)
7.10
22.24
9.78
5.28
9.54
3.94
9.13
Marginal,
Submarginal and
Soil Bank Land
Total
(1,000 Kg)
27.76
4.10
132.78
14.87
369.68
56.30
605.49
Average
per Hectare
(Kg)
7.35
29.85
11.00
4.64
9.00
3.98
8.13
             *Includes marginal and Submarginal Soil Bank land.

-------
TABLE  15.  ESTIMATED PHOSPHORUS  LOSS  FROM NEW CROPLAND ON MARGINAL,  STJBMARGINAL, AND SOIL
                                    BANK LAND  IN GEORGIA,  1976
Marginal and
Submarginal Land*
Land Resource
Area

Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
State Total
Cultivatable
Soil Bank Land
Marginal ,
Submarginal and
Soil Bank Land
Average Average Average
Total per Hectare Total per Hectare Total per Hectare
(1,000 Kg)
13.05
2.15
84,13
14.79
186.96
8.66
309.74
(Kg)
25.21
61.49
27,42
7.09
10.00
0.66
8.25
(1,000 Kg)
56.48
2.42
188.57
7.53
216.19
1.72
472.85
(Kg)
17.32
23.51
20.94
6.70
9.67
1.62
12.81
(1,000 Kg)
69.52
4.57
272.64
22.32
403.16
10.37
782.58
(Kg)
18.40
33.21
22.59
6.96
9.82
0.73
10.52
*Includes marginal and stiBmarginal  Soil Bank  land.

-------
            TABLE 16.  ECONOMIC PESTS OF MAJOR CONCERN FOR CORN AND SOYBEANS IN GEORGIA
Crop
Weeds
Insects
                                                                        Diseases
Corn
Soybeans
Annual grasses or grass-like weeds
such as crabgrass, goosegrass, Texas
panicum, and sandspur.  Perennial
grasses such as nutsedge and Johnson-
grass, both small-seeded and large^
seeded broadleaf weeds such as lambs-
quarters , pigweeds, cocklebur, ragweed,
sicklepod, jimpsonweed, tropic croton,
Florida beggarweed and morning glories.

Same as above
                             Fall Armyworm
                             Corn Earworm
                             Cutworm
                             Common Stalk Borer
                             Corn Leaf Aphid
                    Seed decay
                    Seedling blight
                    Seedling root rot
                             Mexican Bean Beetle
                             Velvet Bean Caterpillar
                             Green Cloverworm
                             Corn Earworms,  Loopers
                             Stink Bugs
                             Lesser Cornstalk Borer
                    Seed decay
                    Leaf spots
                    Wildfire
                    Downy mildew
                    Bacterial  blight
                    Root-knot

-------
TABLE 17.  SELECTED CHARACTERISTICS RELATED TO POTENTIAL ENVIRONMENTAL IMPACT OF PESTICIDES USED
                              ON CORN AND SOYBEANS IN GEORGIA, 1976
Average
Application Rate
Pesticide
Acephate
Alanap
Ametryne
Atrazine
Azodrin (Monocrotophos)
Benomyl
Bentazon
B. Thuringeinsis
Cap tan t
Carbaryl
Chloroxuron
Copper Hydroxide
Cyanazine
Dalapon
Dicamba
Di-Syston
2, 4-D
2, 4-DB
DBCP
DDT
DMBP (Dinoseb)
Endrin
AI(mg)/
Hectare
0.06
0.27
0.03
0.63
0.00
0.04
0.04
0.04
0.10
0.72
0.06
0.12
0.13
<0.01
0.03
0.01
0.17
0.01
0.54
0.00
0.15
0.00
Rating*
1
2
1
3
0
1
1
1
1
3
1
2
2
0
1
0
2
0
3
0
2
0
Toxicity to
Fish
LC50 #
Mg/1

>180
=50
12.6
7.0
0.5
190

0.13
1.0
20
=15
4.9
>100
35
.040
4.5->50
4.0

.002
.40
.0002
Rating**

1
1
2
2
2
1

3
2
2
2
2
1
1
3
2
2

3
3
3
Approximate
Persistence
in Soil
Days

20-60
30-90
300-500
20
15
200-400

20
10
300-400
<100
200-400
15-30
60
30
30
30

>1500
15-30
>1000
Rating* #

1
1
3
0
0
2

0
0
3
1
2
0
1
1
1
1

4
0
4
Index of
Potential
Impact

2
1
18
0
0
2

0
0
6
4
8
0
1
0
4
0

0
0
0
                                                                                      (continued)

-------
TABLE 17.  (CONTINUED)
Average
Application Rate
Pesticide
E-Parathion
EPTC
Lasso (Alactilor) '"*"

Linuron^

Me thorny 1
M-Parathion
Metribuzin
Oryzalin
Paraquat
Phorate
Simazine
Sulfur
Sutan (Butylate)
Thiram''"
Toxaphene
Trichlorofon
Trifluralin
Vernam (Vernolate)
AI(mg)/
Hectare Rating*
0.01
0.09
0.13
0.13
0.01
0.01
0.37
0.54
0.02
0.01
<0.01
0.00
<0.01
0.06
1.39
0.10
1.08
0.00
0.34
0.01
0
1
2

1

2
3
1
0
0
0
0
1
3
1
3
0
2
0
Toxicity to
Fish
LC50"
Approximate
Persistence
in Soil
Mg/1 Rating** Days Rating**
.047
19.0
2.3

16

= .9
1.9
>100
*50
400
.0050
5.0
200
4.2
0.75
.003
.16
~ 1
9.6
3
2
2

2

2
2
1
1
0
3
2
0
2
2
3
3
3
2
7-11
30
40-70

120

30
7-11
150-200
90-180
>500
15
200-400
=60
40-80
15
>1500
30
120-180
50
0
1
1

2

1
0
2
2
3
0
2
1
1
0
4
1
2
1
Index of
Potential
Impact
0
2
4

4

4
0
2
0
0
0
0
0
6
0
36
0
12
0
                                             (continued)

-------
                                      TABLE 17.  (CONTINUED)
 *Rating was as follows:  0 = .01 or less except for linuron because used on both corn and soy-
  beans, 1 = .02 to .10, 2 = .11 to .50, and 3 =  >.50.

 #Test commonly conducted for 24 to 96 hours on bluegills, trout or other fish.

**Rating was as follows:  0 = >20Q (relatively nontoxic), 1 = 20-200 (slightly toxic), 2 = .5-20
  (moderately toxic), and 3 = <.50 (highly toxic).

##Rating was as follows;  0 = <30 days (nonpersistent), 1 = 30-90 (slight persistence), 2 = 90-300
  (low persistence), 3 = 300-540 (moderate persistence), and 4 = >540 (persistent).

 tUsed as seed treatment only,

ttUsed on both corn and soybeans.

-------
the index, which  is  reported in Table 17,  is. zero for 20 out  of  37 pesticides,
Pesticides with an index value of zero were considered to have essentially no
potential environmental impact and were therefore excluded from  further
analysis.

     Pesticides with a nonzero index for potential environmental impact were
considered in greater detail.   Toxaphene had the maximum rating  for all three
factors, giving it the highest index value.   Atrazine had the second highest
index because of  the relatively large quantity used and slightly toxic and
moderately persistent characteristics.  Other characteristics must also be
considered in assessing potential environmental impact.   These factors
include mode of application, mode of transport, and most importantly, the
biomagnification  potential.   The 17 pesticides having a nonzero  index value
are characterized according  to these factors in Table 18.   The only pesticide
used in Georgia in 1976 on corn and soybeans that had a possibility of bio-
accumulation was  toxaphene,  a chlorinated hydrocarbon.   The biomagnification
factor for toxaphene was 1,  when referenced to DDT, which had a  biomagnifica-
tion factor of 2.

     Application  rates for pesticides with a potential for environmental
impact were aggregated according to quantity of active ingredients.  Using
this approach, corn  and soybeans had approximately the same application rates
per hectare.  Taking hectares of corn and soybeans on new cropland into con-
sideration, aggregate amounts of pesticides with a potential  environmental
impact used on marginal, submarginal, and Soil Bank land are  reported by LRA
in Table 19.  The Southern Coastal Plain had 110 metric tons  of  active in-
gredients of pesticides with a potential for environmental impact.  The
Piedmont and Atlantic Coast  Flatwoods each had over 30 metric tons of active
ingredients of these pesticides.

     It would be  difficult to assess the impact of any other  pesticide except
in isolated instances related to probability of rainfall shortly  after appli-
cation.  The average monthly precipitation in Georgia is about 10.7 centi-
meters (Figure 8) with a minimum of about 6.9 centimeters  in  October and a
maximum of about  14.2 centimeters in March and again in July.  Rainfall pat-
terns vary somewhat  throughout the climatic regions but the peaks occur at
the same time.  Although monthly precpitation shows two distinct  peaks, water
resources data on river flow (Figure 9) give only one peak period of dis-
charge.  Rivers located in North, Central and South Georgia all  show the same
trend, with peak  discharge rates in February, March, and April.  Much higher
evapotranspiration rates and groundwater recharge during the  summer months
offer some explanation of why precipitation in July has less  influence on
river flow than that in the  early spring months.

     Less vegetative cover,  especially where row crops were harvested the
previous year, would also have an important influence on runoff  pattern and
river flow during early spring months also.   Because normal planting dates
for corn and soybeans (Figure 8) correspond to minimum periods of runoff and
discharge from rivers,  rapidly degradable pesticides applied  at  or near
planting would have  little chance of.polluting major rivers.  Persistent
pesticides applied to the soil surface at any time, however,  could possibly


                                       65

-------
TABLE 18.  APPLICATION INFOBMATION, TRANSPORT MODE,  AND BIOMAGNIFICATION OF SELECTED PESTICIDES
                         CAPABLE OF ENVIRONMENTAL IMPACT IN GEORGIA,  1976
Crop
Pesticide Treated
Alanap
Ametryne

Atrazine

Bent az on

Chloroxuron

Copper Hydroxide
Cyanazine
Dicamba

2, 4-D

EPTC

Lasso (Alachlor)
Linuron

Methomyl

Metribuzin
Sutan (Butylate)
Toxaphene

Trifluralin
S
C

C

S

S

S
C
C

C

C

CS
CS

S

S
C
S

S
Predominant
Chemical Comments on Mode and Transport
Group^ Time of Application** Mode#*
H
H

H

H

H

F
H
H

H

H

H
H

I

H
H
I

H
Soil surface at planting
Directed spray 3-4 weeks after
planting
Soil surface at planting (may be
incorporated)
Broadcast spray 4-5 weeks after
planting
Directed spray 2-3 weeks after
planting
Incorporated at planting
Soil surface at planting
Directed spray 5-6 weeks after
planting
Broadcast or directed spray 4-6
weeks after planting
Incorporated 1 week prior to
planting
Soil surface at planting
Directed spray 3-4 weeks after
planting
Broadcast spray 4 weeks after planting
until bloom
Soil surface at planting
Incorporated 1 week prior to planting
Broadcast spray 4 weeks after planting
until bloom
Incorporated 1 week prior to planting
W
SW

SW

W

S

W
SW
W

SW

SW

SW
S

u

W
S
S

S
Biomagnif ication
Factor"*"
0
-

0

-

-

0
0
0

0

0

0
0

0

-
0
1

0
(continued )

-------
                                              TABLE 18.   (CONTINUED)
ON
        *_C denotes  chemicals  used on corn, ^ denotes  chemicals  used on soybeans,  and CS denotes chemicals
        used  on both corn and  soybeans.

        #Ą_ denotes  fungicides,  II denotes  herbicides,  and ^ denotes  insecticides.

       **Based on recommendations from Georgia Agricultural and Entomology Extension Service Bulletins.

       ##Where movement of pesticides in  runoff from  treated fields occurs, ^ denotes those chemicals that
        will  most  likely move  primarily  with sediment, W  denotes those that will most likely move pri-
        marily with the water, and SW denotes those  that  will  likely move in appreciable proportion with
        both  sediment and water.

        "tRosmarie von Rumker, et aJL.  Production,  Distribution, Use and Environmental Impact Potential
         of  Selected Pesticides, EPA 540/1-74-001.  Environmental Protection Agency,  Washington,  D.C.,
         1974.

-------
            TABLE 19.  AMOUNT OF PESTICIDES WITH POTENTIAL ENVIRONMENTAL IMPACT APPLIED  TO NEW CROPLAND
                ON MARGINAL, SUBMARGimL AND SOIL BANK LAND BY LAND RESOURCE AREA IN GEORGIA,  1976
00
Land Resource
Area
Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
State Total

Marginal and
Submarginal Land
1.39
0.09
8.26
5.61
50.28
35.27
100.90
Active Ingredients
Cultivatable
Soil Bank Land
CMetric Tons)
8.77
0.28
24.21
3.03
60.20
2.84
99.33

Marginal,
Submarginal ,
and Soil Bank Land
10.16
0.37
32.47
8.64
110.48
38.11
200.23

-------
reach major rivers in Georgia during spring  and summer months.  Pesticides
applied to soybeans during the growing season would he most susceptible  to
storm runoff during July.  Pesticides applied to either corn or soybeans
with, carry-over potential to spring of the following year would be susceptible
to runoff and transport within major rivers  of Georgia.
         JAN  FEB  MAR APR  MAY   JUN  OUL  AU6  SEP  OCT  NOV  DEC

                                   MONTH


             Figure 8.   Average monthly precipitation for Georgia.
                                      69

-------
                  Water Resources for Georgia
Ł«*U
ion
lOU
ior\
IZU


60

n








































































Oostanaulo River at Resaca
Drainage area, 4170 sq km






nrn r-n-n-i
        JAN FEB MAR APR MAY JUN  JUL AUG  SEP OCT NOV DEC  YEAR
8
UJ
co |2O

-------
                                   SECTION 7

                     ESTIMATED ECONOMIC IMPACT ON GEORGIA
     The 74.5  thousand  hectares of new cropland on marginal, submarginal, and
Soil Bank land produced $23.6 million worth of agricultural products in 1976
(Table 20).  Half  of  this  production was in the Southern  Coastal Plain, 19
percent in the Atlantic Coast Flatwoods, and 12 percent was in the Piedmont.
The following  sections  analyze the impact of this  production on net farm
income and Georgia's  economy.
NET FARM INCOME

     Production  from new cropland  on marginal,  submarginal, and Soil Bank
land resulted in a net  increase of $4.5 million in net farm income annually
(Table 21).  This figure is  referred to as  a net increase because it accounts
for the income from  woodland and pastureland that is  foregone to produce
crops.  In other words,  it shows how much additional  net farm income was
generated from crops over the income that would have  been generated if land
use had not changed.  Although the net income is an annual flow, it will
fluctuate from year  to year  depending on such factors as prices received,
prices paid, and average yield per hectare.

     The increased net  income per  hectare for all marginal and submarginal
land averaged $26 compared to $80  for Soil  Bank land  classified as suitable
for production (Table 21).   The smallest difference in increased net income
between these two classifications  occurred  in the Atlantic Coast Flatwoods.
For all marginal, submarginal, and Soil Bank land, average net income per
hectare increased $61 in crop production over other uses.

     Aggregate net farm  income increased $2.4 million in the Southern
Coastal Plain as a result of placing marginal,  submarginal, and Soil Bank
land in crop production.  Net farm income increased $1 million in the Pied-
mont, $0.7 million in the Atlantic Coast Flatwoods, and $0.4 million in the
Appalachian Valley.


AGGREGATE ECONOMIC EFFECTS

     Increased agricultural  production and  farm income affect not only agri-
culture but also many other  sectors of a state's economy.  The most direct
effect is on firms that  supply inputs to agriculture  and firms that process
agricultural products.   Income and employment in these agribusiness firms


                                      71

-------
                  TABLE 20.   ESTIMATED PRODUCTION FROM NEW CROPLAND  ON MARGINAL,  SUBMARGINAL,
                                      AND SOIL BANK LAND IN GEORGIA,  1976
K>
Marginal and
Submarginal Land*
Average
Land Resource Area Total per Hectare
(1,000 Dollars) (Dollars)
Appalachian Valley and
Ridges 143.76 277.67
Mountain 8.63 246.98
Piedmont 765.85 249.40
Sand Hills 400.65 191.92
Southern Coastal
Plain 4,210.76 225.16
Atlantic Coast
Flatwoods 4,055.12 309.12
State Total 9,584.76 255.40
Marginal,
Cultivatable Submarginal and
Soil Bank Land Soil Bank Land
Average Average
Total per Hectare Total per Hectare
(1,000 Dollars) (Dollars) (1,000 Dollars)
1,183.73 362.82 1,327.49
38.19 370.87 46.82
3,104.02 344.73 3,869.87
355.68 316.19 756.33
8,968.08 400.57 13,178.84
388.62 367.98 4,443.74
14,038.32 380.04 23,623.08
(Dollars)
351.15
340.28
320.49
235.44
320.74
313.50
317.23
    *Includes marginal  and  Submarginal Soil Bank land.

-------
TABLE 21.  ESTIMATED INCREASED NET FARM INCOME FROM NEW CROPLAND  ON MARGINAL,  SUBMARGINAL, AND
                                SOIL BANK LAND IN GEORGIA,  1976
Marginal and
Submarginal Land*
Land Resource
Area Total
Average
per Hectare
Marginal,
Cultivatable Submarginal and
Soil Bank Land Soil Bank Land
Total
Average
per Hectare Total
Average
per Hectare

Appalachian Valley and
Ridges 35,074
Mountain -297
Piedmont 114,424
Sand Hills -7,000
Southern Coastal Plain 298,219
Atlantic Coast Flatwoods 591,579
State Total 1,031,999
68
-9
37
-3
16
45
26
— — — — — \LHjjLj-ai. a j
367,346 113 402,420
5,284
869,684
69,543
2,119,807
65,687
3,497,351
51 4,987
97 984,108
62 62,543
95 2,418,026
62 657,266
80 4,529,350
106
36
81
19
59
46
61
     ^Includes marginal and Submarginal Soil Bank land.

-------
are directly linked to agricultural production.  The effects of increased
production do not, however, stop here, as agribusiness firms purchase from
other industries and workers in agriculture and agribusiness spend their
money for goods and services produced by other industries.

     Analytical procedures to quantify the interindustry flows between agri-
culture and nonfarm sectors are complex.  A model has been developed in this
study, however, to quantify the effect of a change in agricultural production
on important economic variables.  Development and interpretation of the
model are presented in Appendix B.

     In addition to explaining economic flows, the interindustry model can
be used to calculate the impact of increased agricultural production.  The
actual Georgia situation with respect to all independent variables in the
interindustry model was taken as the base situation.   Then the effect of
placing the 74.5 thousand hectares of marginal, submarginal, and Soil Bank
land into crop production was evaluated.  It was assumed that (a) production
would increase by $23.6 million, (b) farm earnings -  net farm income and
farm labor earnings - would increase by $6.7 million, and (c) farm purchases
would increase by $15.1 million.  The effect of these changes were simulated
using the interindustry model from Appendix B.  Results of this simulation
analysis indicate that the increased crop production  on marginal, submarginal
and Soil Bank land would increase personal income in  the state by $9.1 mil-
lion.  One-fourth of the increased personal income resulting from increased
crop production accrues to the nonfarm sector.
                                      74

-------
                                 BIBLIOGRAPHY


Alonso, William.  1964.   Location and Land Use:   Towards a General  Theory of
     Land Rent.  Harvard  University Press, Cambridge,  Mass.

Alt, Klaus F.  1976.  An  Economic Analysis of Field Crop Production, Insecti-
     cide Use and Soil  Erosion in a Subbasin of  the Iowa River.  Ph.D. Dis-
     sertation.  Iowa State University,  Ames, Iowa.

Alt, K. F., J. A. Miranowski,  and E. 0.  Heady.   1979.   Social  Costs and
     Effectiveness  of Alternative Nonpoint Pollution Control Practices.  In:
     Best Management Practices for Agriculture and Silviculture.  Proceedings
     of the 1978 Cornell  Agricultural Waste Management Conference.  Ann
     Arbor Science  Publishers, Inc., Ann Arbor,  Michigan,   pp. 321-327.

Anderson, Clayton Qswal,   1978.   Simulation of the Peanut Production Area
     Economy in Georgia:   Implications for Changes in the Peanut Program.
     Masters Thesis.  University of Georgia, Athens,  Georgia.

Armstrong, D. E., and J.  G, Konrad.   1974.  Nonbiological Degradation of
     Pesticides.  In:   Pesticides in Soil and Water.   W.  D. Guenzi  (ed.).
     Soil Scl. Soc. Amer.,  Inc., Madison, Wisconsin.

Austin, M. E.  1965.  Land Resource Regions and  Major Land Resource Areas of
     the United States.   Agricultural Handbook No.  296.   Soil  Conservation
     Service, U.S.  Department  of Agriculture, Washington,  D.C.

Baker, J. L., H. P. Johnson, M.  A. Borcherding,  and W.  R.  Payne.  1979.
     Nutrient and Pesticide Movement from Field  to Stream:  A  Field Study.
     In:  Best Management Practices for Agriculture and Silviculture.  Pro-
     ceedings of the 1978 Cornell Agricultural Waste Management Conference.
     Ann Arbor Science  Publishers, Inc., Ann Arbor, Michigan,  pp.  213-245.

Baldwin, F. L. , P.  W. Santlemann, and J. M. Davidson.   1975.   Prometryn Move-
     ment Across and Through the Soil.  Weed Science.   23:285-288.

Barkley, Paul W. and David W.  Seckler.  1972. Economic Growth and  Environ-
     mental Decay.  Harcourt Brace Jovanovich, Inc.

Barlowe, Raleigh.   1978.   Land Resource Economics.   3rd Edition.  Prentice-
     Hall, Inc., Englewood Cliffs, New Jersey.

Barnett, A. P., E.  W. Hauser,  A. W.  White and J. H. Holladay.  1967.  Loss  of
     2, 4-D in Washoff  from Cultivated Land.  Weeds.   15:133-137.


                                      75

-------
Baumol, William J. and Wallace E. Dates.  1979.  Economics, Environmental
     Policy and the Quality of Life.  Prentice-Hall, Inc., Englewood  Cliffs,
     New Jersey.

Bishop, B. C.  1976.  Landsat Looks, at Hometown Earth.  National Geographic.,
     150:140-147.

Bradley, J. R., Jr., T. J. Sheets, and M. D. Jackson.  1972.  DDT and Toxa-
     phene Movement in Surface Water from Cotton Plots.  J. Environmental
     Quality.  1:102-105.

Brandow, G. E.  1977.  Policy for Commercial Agriculture, 1945-71.  In:   A
     Survey of Agricultural Economics Literature, Volume 1, Traditional
     Fields of Agricultural Economics, 1940s to 1970s.  Lee R. Martin (ed.).
     University of Minnesota Press, Minneapolis,  pp. 209-292.

Briggs, G. G,  1973.  A Simple Relation Between Soil Adsorption of Organic
     Chemicals and Their Octanol/Water Partition Coefficients,  In:   Proceed-
     ings of the 7th British. Insecticide and Fungicide Conference.

Buchanan, J. M. and W. C. Stubblebine,  1962.  Externality.  Economica.   N.S.
     Vol. 29, No. 116.

Chen, Dean T..  1977.  The Wharton Agricultural Model;  Structure, Specifica-
     tion, and Some Simulation Results.  Amer, J, Agr. Econ.  59[107-116.

Chiou, C. T., V. H. Freed, D. W. Schmedding, and R. L. Kohnert.  1977.  Par-
     tition Coefficient and Biomagnification of Selected Organic Chemicals.
     Environmental Science Tech,  11:475-478.

Cicchetti, C. J. and V. K, Smith.  1976.  The Costs of Congestion:  An Econo-
     metric Analysis of Wilderness Recreation.  Ballinger Publishing  Co.,
     Cambridge, Mass.

Cromarty, William A.  1959.  An Econometric Model for United States Agricul-
     ture.  J. Amer. Stat. Assoc.  54:556-574.

Crosby, D. G.  1970.  The Nonbiological Degradation of Pesticides in  Soils.
     In:  Pesticides in the Soil;  Ecology, Degradation, and Movement.
     Internal Symposium on Pesticides in the Soil, Michigan State Univ.,
     East Lansing.

Currie, John Martin, John A. Murphy, and Andrew Schmitz,  1971.  The  Concept
     of Economic Surplus and Its Use in Economic Analysis..  The Economic
     Journal.  81;741-799.

Czamanski, Stanislaw and E. E. Malizia.  1969.  Applicability and Limitations
     in the Use of National Input-Output Tables for Regional Studies.  The
     Regional Science Association Papers.  23;65-83.
                                      76

-------
Davidson, J. M. , G,  H.  Brasewitz, D. R. Baker, and A. L. Wood.  1975   Use of
     Soil Parameters for Describing Pesticide Movement Through Soils.   EPA
     66Q/2-75-OQ7, U.S.  Environmental Protection Agency, Washington, B.C.

Dillion, John L.   1977.   The Analysis of Response in Crop and Livestock Pro-
     duction.   2nd Edition.   Pergamon Press, N.Y.

Doeksen, Gerald A. and Charles H. Little.  1969.  An Analysis of Oklahoma's
     Economy by Districts Using Input-Output Techniques.  So. J. Agr.  Econ.
     pp. 27-36.

Eckstein, Otto.  1958.   Water Resource Development.   Harvard University Press,
     Cambridge, Mass.

Edwards, W. M., and  B.  L. Glass.  1971.  Methoxychlor and 2,4,5-T in Lysimeter
     Percolation and Runoff Water.  Bui. Environmental Contimination Toxicol.
     6:81-84.

Evans, Michael  K.  1969.  An Agricultural Sub-model for the United States
     Economy.   In:   Essays in Industrial Economics II.  L.  Klein (ed.).
     University of Pennsylvania,  pp. 63-145.

Fox, Karl A.   1965.   A Sub-model of the Agricultural Sector.   In;   The  Brook-
     ings Quarterly  Econometric Model of the United States.   James S. Deusen-
     berry, Garry  Fromm, Lawrence R. Klein, and Edwin Kuh,  (eds,)-   Rand
     McNally &  Co.,  Chicago,  pp. 409-461.

Freeman, A. Myrick III.   1979.  The Benefits of Environmental Improvement:
     Theory and Practice.  John Hopkins University Press,  Baltimore,
     Maryland.

Georgia Forestry Commission.  1976.  A Report on the Current Status of  the
     Pine Plantations Established in the State of Georgia Under  the 1956 to
     1960 Soil  Bank, Conservation Reserve Program,  Macon,  Georgia.

Guntermann, K.  L., Ming T. Lee, and E. R. Swanson.  1975.   The Off-Site Sedi-
     ment Damage Function in Selected Illinois Watersheds.   Journal of  Soil
     and Water  Conservation.  30:219-224.

Hall, J. K.  1974.   Erosional Losses of S-triazine Herbicides.   J.  Environ-
     mental Quality.  3:174-180.

Hartung, R.  1975.   Accumulation of Chemicals in the Hydrosphere.   In:
     Environmental Dynamics of Pesticides.   R. R. Hague and V. H,  Freed,
     (eds.).  Plenum Press,  New York, N.Y,   pp. 217-273,

Heady, E. 0, and W.  Candler.  1964,  Linear Programming Methods.  Fourth
     Printing.  Iowa State Press, Ames, Iowa.
                                      77

-------
Hockensmith, Roy D.  1948.  Classification of Land According  to  Its  Capa-
     bility as a Basis for a Soil Conservation Program,  Reprinted from Pro-
     ceedings of the Inter-American Conference on Conservation of Renewable
     Natural Resources.  Denver, Colorado.

Holt, R. F.  1969.  Runoff and Sediment as Nutrient Sources.  Presented at
     1969 Annual Meeting of Minnesota Chapter Soil Conservation  Society of
     America.  Bulletin No. 13.  University of Minnesota, Minneapolis,

Homer, Gerald L.  1975.  Internalizing Agricultural Nitrogen Pollution
     Externalities:  A Case Study.  Amer. J. Agr. Econ.  57:346-354.

Huang, J. C.  1971.  Effect of Selected Factors on Pesticide  Sorption and
     Desorption in the Aquatic System.  J. Water Pollution Control Federa-
     tion.  43:1, 739.

Jaworski, N. and L. J. Hetling,  1970,  Relative Contributions of Nutrients
     to the Potomac River Basin from Various Sources.  In:  Relationship of
     Agriculture to Soil and Water Pollution.  Conference of  Agricultural
     Waste Management, Cornell University, Ithaca, New York.

Jenkins, R,, T. Eichers, P. Andrilenas, and A. Fox.  1969.  Pesticide Appli-
     cation Equipment Owned by Farmers, 48 States.  Agr, Econ. Report No. 161.
     USDA, ERS.  pp. 1-5.

Joncker, Marc.  1974.  An Economic Model for the Heart of Georgia Area  Plan-
     ning and Development Commission,  Masters Thesis.  University of Georgia,
     Athens, Georgia.

Kaufman, D. D.  1974.  Degradation of Pesticides by Soil Microorganisms.  In:
     Pesticides in Soil and Water.  W. D. Guenzi, (ed.).  Soil Sci, Soc,  Amer.,
     Inc.  Madison, Wisconsin,

Kenaga, E. E.  1975.  Partitioning and Uptake of Pesticides in Biological
     Systems,  In:  Environmental Dynamics of Pesticides,  R. Hague  and V.  H.
     Freed, (eds.).  Plenum Press, New York, N.Y.  pp. 217-273.

Klemas, V. and D. Bartlett.  1974.  Inventories of Delaware's Coastal Vegeta-
     tion and Land-Use Utilizing Digital Processing of ERTS-1 Imagery.   In:
     Proceedings of the 9th International Symposium on Remote Sensing of
     Environment.  Ann Arbor, Michigan.

Kneese, Allen V. and Blair T. Bower.  1968,  Managing Water Quality;
     Economics, Technology, Institutions.  The John Hopkins Press, Baltimore,,
     Maryland.

Knisel, W. G., Jr., and R. W. Baird.  1969.  Runoff Volume Prediction Using
     Daily Climatic Data.  Water Resources Research.  5:84-89.

Krutilla, John V. and Anthony C. Fisher.  1975.  The Economics of Natural
     Environments.  John Hopkins University Press, Baltimore, Maryland.
                                      78

-------
Langham, M. R., J.  C. Headley,  and Ą.  F. Edwards.  1972.  Agricultural Pesti-
     cides:  Productivity  and Externalities.  In:  Environmental Quality
     Analysis.  The John Hopkins Press, Baltimore, Maryland.

Langley, L. H.  1972.   An  Input-Output Analysis of Georgia's  Agribusiness
     Industry.  Masters Thesis.   University of Georgia,  Athens,  Georgia.

Leonard, R. A., G.  W. Bailey, and R. R. Swank, Jr.  1976.   Transport,  Detoxi-
     fication, Fate,  and Effects of Pesticides in Soil and Water Environ-
     ments.  Land Application of Waste Materials,  Soil Conservation Society
     of America,  pp. 48-78.

Liu, Peng-Li.  1975.  An Aggregate Economic Analysis of  the Coosa Valley Area
     of Georgia, 1972.   Masters Thesis.  University of Georgia,  Athens,
     Georgia.

MacDonald, R.  B. , F.  G. Hall, and R. B. Erb.  1975.  The Large Area  Crop
     Inventory Experiment  (LACIE) - An Assessment After One Year of  Operation.
     In:  Proceedings of the  10th International Symposium on  Remote  Sensing
     of Environment.  Ann  Arbor, Michigan.

McArthur, W. C.  1971.   Selected U.S.  Crop  Budgets, Yields, Inputs,  and Vari-
     able Costs.  Vol.  I., Southeast Region.  Economic Research  Service,
     U.S. Department of Agriculture, Washington, D.C.

McCarl, T. A.  1971.  Quality and Quantity  of Surface  Runoff  from a  Cropland
     Area in South  Dakota  During 1970.  M.S. Thesis.   South Dakota State
     University, Brookings,  South Dakota.

McConnell, Kenneth  E.   1977.   Congestion and Willingness to Pay:   A  Study of
     Beach Use.  Land Economics.  53:185-195.

McElroy, A. D. , S.  Y. Chiu, J.  W. Nebgen, A. Aleti, and  F.  W.  Bennett.   1976.
     Loading Functions  for Assessment  of Water Pollution from Nonpoint
     Sources.  EPA-600/2-77-151, Office of  Research and  Development, U.S.
     Environmental  Protection Agency,  Washington, D.C.

McKean, Roland.  1968.   Public Spending. McGraw-Hill,  N.Y, N.Y.

Meta Systems,  Inc.   1979.   Costs and Water  Quality Impacts of Reducing Agri-
     cultural  Nonpoint  Source Pollution; An Analysis  Methodology.   EPA-
     600/5-79-009,  U.S. Environmental  Protection Agency, Athens,  Georgia.

Miernyk, W. H.  1967.   The Elements of Input-Output Analysis.  Fifth Printing.
     Random House,  New  York,  N.Y.

Miller, Bill R. and Peng-Li Liu.  1975.  Regional Input-Output Models
     Adjusted  by Import-Export Survey  Data.   So. J. Agr. Econ.   7:61-68.

Morrison, W. I. and P.  Smith.  19.74.  Nonsurvey Input-Output  Techniques  at the
     Small Area Level:   An Evaluation.  J.  Regional Science.   14:1-4.
                                       79

-------
National Research Council, Committee on Nitrate Accumulation,   1972.   Accumu-
     lation of Nitrate.  National Academy of Sciences, Washington, D..C.

National Research. Council, Panel on Nitrates.  1978.  Nitrates:  An  Environ-
     mental Assessment.  National Academy of Sciences, Washington, B.C.

National Weather Service.  Annual.  Climatological Data - Georgia,   National
     Oceanic and Atmospheric Administration, Washington, B.C.

Nicholson, H. P.  1975.  Agricultural Chemicals and Water Quality.   In;
     Water Resources Utilization and Conservation in the Environment.  M.  C.
     Blount, (ed.).  Taylor County Printing Co., Reynolds, Georgia,  pp.  214-
     227.

Ogg, Clayton W. , Lee H. Christensen, and Ralph E. Heimlich.  1979.   Economics
     of Water Quality in Agriculture - A Literature Review.  ESCS-58,  ESCS,
     U.S. Department of Agriculture, Washington, D.C,

Omernik, J. M.  1976.  The Influence of Land Use on Stream Nutrient  Levels.
     Economic Research Series, EPA-600/3-76-Q14, Office of Research  and
     Development, U.S. Environmental Protection Agency, Corvallis, Oregon.

Osteen, Craig and Wesley D. Seitz,  1978.  Regional Economic Impacts of
     Policies to Control Erosion and Sedimentation in Illinois and Other  Corn
     Belt States.  Amer. J. Agr. Econ.  60:510-517.

Paxton, Kenneth W.  1977.  Cotton and Soybean Production, Coats and Returns.
     Department of Agricultural Economics and Agribusiness, Louisiana  State
     University, Baton Rouge.

Ritter, W. F., H. P. Johnson, W. G. Lovely, and M. Molnau.  1974.  Atrazine,
     Propachlor and Diazinon Residues on Small Agricultural Watersheds.
     Environmental Science Tech.  8:38-42.

Roehl, J. W.  1962.  Sediment Source Areas, Delivery Ratios, and Influencing
     Morphological Factors.  Commission on Land Erosion.  Publication  No.  59.
     International Association of Scientific Hydrology,  pp. 202-213.

Roop, Joseph M. and Randolph H. Zeitner.  1977.  Agricultural Activity and
     the General Economy:  Some Macromodel Experiments.  Amer. J. Agr. Econ.
     59:117-125.

Schaffer, W. A., E. A, Laurent and E. M. Sutter.  1972.  Introducing the
     Georgia Economic Model,  The Georgia Department of Industry and Trade,
     Atlanta, Georgia.

Scott, R. B. and R. A. Harding.  1975.  Satellite and Airplane Remote  Sensing
     of Natural Resources in the State of Washington.  In:  Proceedings of
     the 10th International Symposium on Remote Sensing of Environment.   Ann
     Arbor, Michigan.
                                      80

-------
Seitz, W. D. , C. Osteen,  and M.  C.  Nelson.   1979.   Economic Impacts  of Poli-
     cies to Control  Erosion and Sedimentation in Illinois and Other Corn-
     Belt States.   In:   Best Management Practices  for Agriculture  and Silvi-
     culture.  Proceedings of the 1978 Cornell Agricultural Waste  Management
     Conference.  Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan.

Siegel, Sidney.  1956.   Non Parametric Statistics  for the Behavioral Sciences.
     McGraw-Hill, N.Y.,  N.Y.

Shulstad, Robert N. and  Herbert H.  Stoevener.   1978.   The Effects  of Mercury
     Contamination  in Pheasants on the Value of Pheasant  Hunting in  Oregon.
     Land Economics.  54:39-49.

Smith, G. E., F. D. Whitaker, and H, G. Heinemann.   1974.   Losses  of Ferti-
     lizer  and Pesticides from Clay Pan Soil.   EPA-660/2-74-068, U,S. Environ-
     mental Protection Agency, Washington,  D.C.

State  Soil  and Water  Conservation Committee of Georgia.   1977.  Manual for
     Erosion and Sediment Control in Georgia.   Athens, Georgia.

Stevens, Joe B.  1966.   Recreation Benefits from Water Pollution,Control.
     Water  Resources  Research.  2:167-182.

Taylor, C.  Robert and Klaus K. Frohberg.  1977. The  Welfare Effects of Ero-
     sion Controls, Banning Pesticides, and Limiting  Fertilizer Applications
     in the Corn Belt.   Amer. J. Agr. Econ.  59:25-35.

U.S. Congress.  1977. Costs of Producing Selected Crops  in the United States-
     1975,  1976 and Projections for 1977.  ERS, U.S.  Dept.  of  Agriculture.
     The Committee  on Agriculture and Forestry, United States  Senate, 95th
     Congress, 1st  Session.

U.S. Department of  Agriculture.   1975.  Perspectives  on Prime  Lands.  Back-
     ground Papers  for Seminar on the Retention of  Prime  Lands.  Sponsored
     by the USDA Committee on Land Use.

U.S. Department of  Agriculture,  Agricultural Research Service.  1976.  Control
     of Water Pollution  from Cropland.  Washington, D.C.

U.S. Department of  Agriculture,  Agricultural Stabilization and Conservation
     Service.  Annual.   Georgia ASCS Annual Reports.  Athens,  Georgia.

U.S. Department of  Agriculture,  Agricultutal Stabilization and Conservation
     Service.  1970.  Final Report, Conservation Reserve  Program:  Summary of
     Accomplishments, 1956-1972.  Washington,  D.C.

U.S. Department of  Agriculture,  Agricultural Stabilization and Conservation
     Service.  1976.  Farm Commodity and Related Programs.   Agriculture Hand-
     book No. 345.  Washington,  D.C.

U.S. Department of  Agriculture,  ESCS.  1976.  Cash Rents.   Unpublished Data.
     Washing ton, D.C.

                                       81

-------
U.S. Department of Agriculture, Economic Research  Service,   1971.   Pesticide
     and General Farm Survey.  Washington, D.C,

U.S. Department of Agriculture, Economic Research  Service.   1975.   Commodity
     Budgets.  Prepared by Firm Enterprise Data  System,  Commodity  Economics
     Division, ERS in cooperation with Oklahoma  State  University.

U.S. Department of Agriculture, Economic Research  Service.   1976,   State
     Farm Income Statistics,  Washington, D.C,

U.S. Department of Agriculture, Soil Conservation  Service,   1959*   Basic
     Watershed Map of Georgia.  Athens, Georgia.

U.S. Department of Agriculture, Soil Conservation  Service..   19172,   Compila-
     tion of Soils Data in Georgia.  Washington, D.C,

U.S. Department of Agriculture, Soil Conservation  Service.   1975.   Urban
     Hydrology for Small Watersheds.  Technical  Release  No.  55,  adapted for
     use in Georgia.  Washington, D.C.

U.S. Department of Agriculture, Soil Conservation  Service.   1970.   Georgia
     Conservation Needs Inventory, 1967.  Washington,  D.C.

U.S. Department of Agriculture, Soil Conservation  Service.   1977.   Soil
     Association Map of Georgia.  Athens, Georgia.

U.S. Department of Agriculture, Statistical Reporting  Service.   1977.   Agri-
     cultural Prices.  Annual Summary 1976.  Washington, D,C.

U.S. Department of Agriculture, Statistical Reporting  Service, Georgia  Crop
     Reporting Service.  1978.  Georgia Agricultural Facts 1969-1977.

U.S. Department of Commerce, Bureau of Economic  Analysis.  1976.   Survey of
     Current Business.  Washington, D.C.

U.S. Department of Commerce, Bureau of the Census. 1972.  1969 Census of
     Agriculture.  Washington, D.C.

U.S. Department of Commerce, Bureau of the Census. 1972.  U.S. Census of
     Population, 1970 and Area of the United States, 1940.   Washington, D.C.

U.S. Department of Commerce, Bureau of the Census. 1976.  Annual Survey of
     Manufacturers, 1975.  M75(A8)-6.  Washington, D.C.

U.S. Department of Commerce, Bureau of the Census.  1973.  U.S.  Census  of
     Retail Trade, 1972.  Area Series, RC72A-1 to  52.  Washington,  D.C.

U.S. Department of Commerce, Bureau of the Census.  1977.  1974  Census  of
     Agriculture.  Washington, D.C.

U.S. Department of Labor, Bureau of Labor Statistics.  1976,  Employment and
     Earnings.   Washington, D.C.

                                      82

-------
U.S. Internal Revenue Service,  1974.  Statistics of Income and Supplemental
     Report Personal Wealth,   Washington., D,C,

von Rumker, Rosmarie, E.  W, Lawless, A, F, Meiners, K. A, Lawrence,  G,  L.
     Kelso, and  Freda Horay,   19174,  Production, Distribution, Use and
     Environmental Impact Potential of Selected Pesticides,  EPA-540/1-74-001,
     U.S. Environmental Protection Agency, Washington, D,C,

Wade, James C. and Earl 0. Heady.  1977.  Controlling Nonpoint Sediment
     Sources with Cropland Management:  A National Economic Assessment.
     Amer. J. Agr. Econ.   59:13-24.

Wadleigh, C. H.   1968.  Wastes in Relation to Agriculture and Forestry.  U.S.
     Department  of Agriculture Pub. No. 1065.  Washington, D.C.

Westin, F. C. and C. J. Frazee.  1976.  Landsat Data, Its Use in a Soil Sur-
     vey Program.  Soil Sci.  Soc. Amer. J.  40:81-89.

White, A. W. , A.  P. Barnett,  B. C, Wright, and J. H. Holladay.  1967.   Atra-
     zine Losses from Fallow Land Caused by Runoff and Erosion.   Environ-
     mental  Science Tech.  1:740-744.

Wiese, A.  1972.   Herbicide Runoff No Great Environmental Threat.   Crops Soil.
     24:23.

Willis, G. H. ,  and R. A.  Hamilton.  1973,  Agricultural Chemicals  in Surface
     Runoff, Ground Water and Soil:  I Endrin.  J. Environmental Quality.
     2:463-466.

Wischmeier, W.  H. and D.  D. Smith.  1965.  Predicting Rainfall-Erosion  Losses
     from Cropland East of the Rocky Mountains—Guide for Selection  of  Prac-
     tices for  Soil and Water Conservation.  Agr. Handbook No. 282.   Washing-
     ton, D.C.

Wise, James  0.   1974.  Selected Crop and Livestock Budgets for the Piedmont
     Area of Georgia.  University of Georgia College of Agriculture  Experi-
     ment Station.  Research Report 188.

Ziemer, R. F. , W. N. Musser,  and I. D. Clifton.  1978,  Agricultural Input
     Use and Potential Water Pollution Associated with Irrigation in Georgia.
     University  of Georgia College of Agriculture Experiment Station.   Research
     Bulletin No. 222.  Athens, Georgia.
                                       83

-------
                                 APPENDIX A

            LANDSAT REMOTE SENSING TO DETEEMINE LAND USE CHANGES
     One of the more important recent technological advances for scientific
research has been the development of a satellite specifically designed to
gather information about the earth's resources.  Two satellites launched by
the National Aeronautics and Space Administration and known as LANDSAT 1 and
2, currently orbit the earth in a systematic fashion every nine days gather-
ing data for virtually every point on the globe.  This new source of data has
important implications for research on natural resource use.

     Researchers interested in monitoring natural resource use have applied
LANDSAT imagery to many problems.  The U.S. Department of Agriculture, the
National Oceanic and Atmospheric Administration, and NASA have created LACIE
(Large Area Crop Inventory Experiment) to help accurately forecast harvests
on a global scale.  After one year of operation, the system implemented to
handle and analyze LANDSAT data demonstrated promise (MacDonald, Hall and Erb,
1975).  In the state of Washington, researchers have found that computer-
aided interpretation of LANDSAT data offers natural resource land planners
an unparalleled opportunity to examine land cover of large areas (Scott and
Harding, 1975).  In another application, Klemas and Bartlett used satellite
imagery to map and inventory the significant ecological communities of
Delaware's coastal zone and concluded that data products were accurate and
could be easily used by researchers, planners, and government officials.

     The purpose of this appendix is to report on the potential for using
LANDSAT remote sensing data to determine changing rural land use patterns.
This study focuses on the practical utility of using satellite imagery as a
primary source of information for determining land cover changes on an inter-
temporal basis.  The research experience reported here should be helpful to
other researchers contemplating implementation of LANDSAT imagery for similar
purposes.

     Central to any decision to implement LANDSAT data, of course, is a
basic understanding of how the satellite works, what kind of information it
collects, and a knowledge of its advantages and limitations.  On board the
spacecraft is a multispectral scanner that records reflected energy from the
earth and channels it through an optical system to detectors sensitive to
four different bands of the spectrum (three of the bands are in the visible
portion of the electromagnetic spectrum and the fourth is in the near infra-
red).   The detectprs measure the light intensities of 0.4 hectare picture
elements called pixels and convert the reflected light from these pixels into
electric voltages.  A digitizer then translates the impulses into computer-


                                     84

-------
digestible number values from zero to 63 and the data are transmitted to
receiving  stations, recorded on tape, and shipped to a central processing
facility at Goddard Space Flight Center.               central processxng

      These LANDSAT data are grouped into LANDSAT scenes.  Each scene is
approximately 185 by 185 kilometers and covers an area of 3.4 million hec-
tares or 34,253 square kilometers.  LANDSAT surveys one scene in 25 seconds,
a_speed that is quite remarkable as one scene contains more than 7.5 million
pixels per band or approximately 30 million pixels for all four bands.   Com-
puter comptabile tapes and 23 by 23 centimeter transparencies that correspond
with  a particular scene can be ordered from the Earth Resources Observation
System (EROS) facility in Sioux Falls, South Dakota.

      In agriculture, researchers can identify several different types of
crops with considerable accuracy in fields as small as 8 hectares.   Success
in  classifying land cover using the LANDSAT data depends to a large degree on
the quality of "ground truth" information.   Ground truth is land cover  that
is  actually identified by fieldwork.  When this information is compared with
the satellite image of a small area, researchers can verify what the pixels
in  the image represent and can then use computer analysis to extrapolate
those results to much larger areas (Bishop,  1976).

      Aside from the almost limitless number  of possibilities and applications
in  resource management, technical advantages of LANDSAT are that the data  are
multispectral, temporal, synoptic, and near  orthographic.   The synoptic
character  of LANDSAT images refers to the fact that LANDSAT imagery covers
enough area to show a wide range of land characteristics.   The near-
orthographic quality of LANDSAT images means that overlays  of controlled
base  maps  fit the images with minimum of distortion.   Thus  they have pro-
perties not previously available in aircraft photography.   Limitations  of
LANDSAT data, however, include lower resolution than aircraft imagery,  atmo-
spheric attenuation and, because data are collected at fixed intervals, cloudy
or  stormy  weather cannot be avoided.   Additional information on the technical
aspects of LANDSAT data is discussed in Westin and  Frazee's  article on  the
use of satellite imagery in a soil survey program.

      Special consideration must be given to  the dates of  the satellite
images used.   Based on NASA training workshops  at Bay St. Louis, Mississippi
and conversations with personnel of the Georgia Department  of Natural Re-
sources and Engineering Experiment Station at Georgia Tech,  mid-summer  was
judged to  be the most favorable time to classify land cover  using LANDSAT
imagery.   During mid-summer most row crops would have sufficient growth to  be
distinguished from bare ground.   With winter scenes,  it would be difficult
to distinguish between idle and cultivated land.  Also,  some land used  for
row crops  might be double cropped with such  close grown  crops as wheat, bar-
ley,  and rye during the winter.   A fall scene would include  some crops  that
had already been harvested;  other crops  would be almost  ready for harvesting.
Imagery with no cloud cover or  haze is,  of course,  the best  imagery to
obtain.  It was believed,  however,  that  imagery  with some cloud cover could
be used if  the clouds were not  over the  area to  be  analyzed.
                                      85

-------
     Ground truth information was collected and delineated for this study  on
U.S. Geological Survey topographic maps (.7.5 min. quad sheets) so that accu-
rate Universal Transverse Mercator (UTM) coordinates for specific fields
could be noted and entered into the computer.  The ground truth information
consisted of several large fields of each major type of land cover.  Experi-
ence with processing LANDSAT data revealed that large fields located near
salient land features such as rivers, lakes, and crossroads provided the best
ground truth.

     Because LANDSAT scenes for two different time periods do not exactly
correspond, the data must be geographically referenced to the UTM coordinate
system.  This procedure allows the data to be superimposed to determine areas
in which land cover changes have taken place between the dates of two LANDSAT
passes.  For this study, researchers located prominent ground features on
black and white photographs that were supplied with each scene, found their
UTM coordinates, and assisted the computer scientist in locating these points
on an image display screen.  After a LANDSAT scene was geographically refer-
enced on the computer,  "training samples"  were identified.   Homogeneous areas
such as centers of lakes, woodland, and corn fields were selected from the
image display screen for input into computer programs that can statistically
determine these various land cover types for other areas in the scene.   Accu-
rate ground truth information is essential in the process of selecting most
of the training samples because the computer would classify a much larger
area based on the samples.
                                     86

-------
                                  APPENDIX B

                INTERINDUSTRY MODEL OF GEORGIA'S STATE ECONOMY


      The purpose of the research reported in this Appendix is to develop  an
 econometric model of interindustry economic flows for state economies.
 Attention is focused on the role of the agricultural sector in the overall
 economy.   Realizing that modeling economic systems is enormously complex, the
 principal economic characteristics of the economy are emphasized.


 RELATED LITERATURE

 National Models

      Previous studies have developed large econometric models of the  agricul-
 ture sector.  Early work includes models by Fox,  Cromarty,  and Evans  but
 these full-size agricultural sector models have not been incorporated into
 large macromodels.   Although these models have been useful  in analysis of
 agriculture in the past, the increasingly important contribution of agricul-
 ture to the general economy as well as macroeconomic effects  on agriculture
 have caused use of such models to be increasingly limited.

      Recent studies have sought to link the agricultural sector to the macro-
 economy.   Chen constructed the Wharton Agricultural Model using quarterly
 data and  22 structural equations.   The model simultaneously determined quar-
 terly equations of supply, demand,  inventory,  and price relations for 17 com-
 modities  or commodity groups.   Annual crop production was projected using
 annual and recursive equations in the annual crop production  block.   The
 income and expenditure accounts for the agriculture sector  were basically
 definitional.   The micro-macro linkage block treats the agricultural  sector
 as other  production sectors of the economy by connecting aggregate demand
 components,  labor and employment,  wages,  and income.   A full  feedback rela-
 tionship  between the agricultural sector  and macromodel results from  these
 connections.   Roop  and Zeitner have also  constructed an agricultural  sector
model compatible with the Wharton macromodel.   They base their  model on the
 income statement as used in the National  Income and Product Accounts.  Nine
behavioral equations were specified to replace  the five agricultural equa-
 tions  in  the original Wharton  model.   The nine  equations include gross farm
product,  farm sector income, net proprietors' farm income,  cash receipts,
investment employment,  implicit sector deflator, wage compensation per man-
year,  and  depreciation expense.
                                      87

-------
     An examination of national models  linking  the  agricultural  sector to
nonfarm sectors may prove useful for formulating  state models.   However,  many
of the relationships  identified in this manner  would be  considerably less
important for area models than for a national model.  Hence,  it  would be
inappropriate to reestimate  these models for state  economies  without con-
siderable reformulation.  Also, the detail in the nonfarm sectors  of these
models is really not  needed  for the purposes of this study.

State Models

     The most frequently used form of macroeconomic model at  the state and
also sub-state levels has been input-output (1-0) analysis.   Input-output
applications have been numerous and the basic techniques for  constructing
input-output tables are described in detail in  Heady and Candler and Miernyk.

     Several 1-0 models have been constructed in  Georgia for  different levels
of aggregation over regions  and over sectors.   These are discussed by
Langley, Schaffer, _e_t ai^. , Joncker, and Liu.  Effects of variation in the
number of sectors in  an 1-0  model have been explored by Doeksen  and  Little
in a simulation study in which they found that multiplier estimates  were
comparable among all  models  regardless of aggregation level.

     The economic flows among sectors have been the subject of several input-
output (1-0) studies  in Georgia.  Langley constructed the first  input-output
study of Georgia in 1969.  His model consisted  of 14 sectors, 12 of  which
were agribusiness sectors.    The other sectors were farming and all-other
industry.  Schaffer,  Laurent, and Sutter constructed the largest Georgia  1-0
table in 1970 containing 50  sectors.

     A principal problem in  all 1-0 studies has been the high cost of  con-
structing transaction tables describing economic  flows in the model.   The
minimum cost of constructing an acceptable 1-0 model for the  state of
Georgia has been estimated to be $100 thousand  (Schaffer, Laurent, and Sutter,
1972).  Even at this  cost the process would consist mainly of adjusting a
national 1-0 table to Georgia conditions.

     Several adjustment techniques have evolved to use 1-0 coefficients in
sub-regions of an area that may have an existing  current model.  Morrison and
Smith in a recent review of  techniques concluded  that a simple location quo-
tient technique is very inexpensive and can be used with ordinarily  available
secondary data sources.  Accuracy of the simple location quotient, (LQ),
method, however, has  been questioned.  Miller and Liu compared estimates
from a location quotient model with location quotient results corrected by
surveys of imports and exports and concluded that there were  significant
errors in the unadjusted LQ method.  Survey techniques for adjusting existing
models have also been devised and recommended by  Czamanski and Malizia.

     As the major importance of 1-0 models is to  estimate the aggregate
impacts on multipliers of exogenous changes occurring in a single sector  of
the economy,  there remains a demand and possibility that aggregate multipli-
ers can be estimated  by econometric techniques applied to available  data.
                                     88

-------
CONCEPTUAL FRAMEWORK

  ^  This survey  of  recent work suggests a possible comparison between the
major economic  flows of  an area and similar transactions  in  an 1-0 table.
Examining some  of the important flows in an 1-0 framework may help to specify
variables and equations  in an appropriate econometric model.

     In a closed  1-0 model, major interactions are found  between agriculture,
basic industry, service  industry and households.   The model, however, is
open^with respect to all exports from the area and all  imports.  As pointed
out in all 1-0  discussions, an important characteristic of the economy is
that the equilibrium flows are simultaneously determined.  Agriculture's
purchases from  basic industry are significant; energy,  fertilizer, insecti-
cides and mechanization  have been at the forefront of agricultural growth.
In state economies,  however, most farmers'  purchases of these items are from
service industries such  as petroleum wholesalers,  fertilizer dealers, and
farm supply  stores in general.  Thus, the principal payments of agriculture
are to service  industry, for imports of primary inputs  and earnings from
farming paid households  in the area.  Basic industry would of course be the
primary buyer of  most agricultural products.   But  again in states, agricul-
tural outputs are likely to move through a service industry buyer and from
there into regional, national and even world markets before eventually being
utilized by  basic industry.

     Local households, as an endogenous sector of  a closed model would be
expected to  purchase very little, if any, goods from local agriculture and
local manufacturing.  Most of their incomes will be spent for goods filtered
through wholesale and retail trade and other service industries.

     Specific transaction data are not generally available to describe input-
output flows.   As indicated by previous studies, our current data systems are
not designed to collect  data of this type.   A great amount of data is col-
lected, however,  that records the results of similar flows described by 1-0.
Thus, conceptually it is possible to specify a set of simultaneous relation-
ships based  on  economic  theory and results of prior 1-0 studies that will
approximately describe aggregate flows and impacts of the same type found by
1-0.
MODEL SPECIFICATION

     'Total  employment,  labor earnings,  and personal  income are thought to be
key performance variables.   An economic model that describes the interrela-
tionships among these variables Is Described below.   The model includes four
behavioral  relationships and four identities.

     Much evidence from prior modeling, particularly 1-0  indicates that
total basic employment  and  total service employment  are determined simul-
taneously    The principal exogenous determinant of employment in basic indus-
tries Is most  likely national demand for output (or  shipments  from these
industries? Although service employment is a function of basic employment,


                                      89

-------
it is also expected to be positively related  to  the  level  of  agricultural
input purchases.  Algebraically, these employment relationships  can be stated
as follows (where X^ represents additional variables  that  may be found to
affect the basic flows):

   Basic employment = f  (Service employment; Value  of  shipments,  X.)   (B-l)
                       _L                                            J

   Service employment = f~  (Basic employment; Farm input purchases,     (B-2)
                        xj)

Determination of the wage structure or even the  average wage  is  far beyond
the scope of this study.  Earnings in the basic  and  service industries are,
therefore, the product of the exogenous industry wage and  endogenous level
of employment.

          Basic earnings =  (Basic wage) (Basic employment)              (B-3)

        Service earnings =  (Service wage) (Service employment)          (B-4)

     Total earnings are the sum of earnings determined  from basic  service
industries and exogenously determined earnings of agriculture and  federal,
state and local government workers.  Farm earnings are  probably  not endo-
genous ly related to other earnings in a state or sub-state economy, but are
probably related almost entirely to farm profits.  Hence,  the total earnings
equation can be expressed as the following identity.

   Total earnings = Basic earnings + Service earnings + Farm earnings   (B-5)
                    + Government earnings

Personal income is endogenously determined as a  function of local  earnings
from employment and from holdings of capital wealth.  Income from  non-local
activities would be small relative to local and probably randomly  associated
with income observations at the state level of aggregation.

           Personal income = f  (Total earnings; Wealth, X.)            (B-6)

Per capita personal income, which will be used to determine retail trade, can
be calcualted directly from total personal income as follows.

             Per capita personal income = Personal income                  }
                                            Population                  v   '
Retail sales are a function of per capita personal income and the  size of
population.   While not all local income is expended locally, a major portion
will be in most areas.  Anderson has shown this hypothesis to hold even for
sub-state levels.

     Retail sales = f, (Per capita personal income; Population,  X.)     (B-8)
                                     90

-------
DATA

     The model  described above was estimated for state economies  for 1975
using cross-sectional data from the 48 contiguous states.   Employment and
earnings by industry group for each state were obtained from  Employment and
Earnings (U.S.  Bureau of Labor Statistics, 1976).  Personal income by state
was reported in Survey of Current Business (U.S. Bureau of  Economic Analysis,
1976).  Farm sales  and farm input purchases were obtained from  State Farm
Income Statistics  (U.S.  Department of Agriculture,  1976).   The  wealth vari-
able used in this study was net worth of top wealth holders by  states
reported in Personal Wealth (U.S. Internal Revenue  Service, 1974).  The rest
of the data including value of shipments in manufacturing,  population, den-
sity, and retail trade was reported by U.S. Bureau  of  the Census.


RESULTS

     The model  was  estimated in logarithms using three-stage  least squares.
The logarithm model was chosen (a) to provide for nonlinearities , (b) to
account for interactions among explanatory variables,  and (c) most impor-
tantly, to measure  marginal rather than average effects of  explanatory vari-
ables.  Marginal impacts are important in this study,  because we  are
interested in measuring the effects of changes in agricultural  production
on the overall  economy.

     When both  the  dependent variable and explanatory  variable  are in loga-
rithms, the regression coefficient is an elasticity; however, the elasticity
can be converted to a marginal effect:

                               ME = b D*/X*                            (B-9)

where

     ME is marginal effect

     b is a regression coefficient and an elasticity

     D* is a specified level of the dependent variable

     X* is a specified level of the explanatory variable

In the following discussion, the elasticities were  converted  to marginal
effects using geometric means for the sample   Because of the structure of
this model, both direct and indirect effects (multipliers)  will be forecasted
by the model.   The  regression coefficients themselves  can be  interpreted as
discussed  below.
The estimates for the parameters of  the model are presented in Table
 S?uden?s "values and  the  level of statistical significance for each
Bl.


                                      91

-------
regression coefficient are also reported in the table.  In general, the model
did a good job in explaining variation in all the behavioral equations, and
almost all of the variables are significant at the 0.01 level.  Only the
wealth variable is not statistically significant; however, the size of the
t-value (1.4) and the expected positive sign indicates that the variable may
reasonably be maintained.  Its lack of significance can probably be attri-
buted to the fact that the variable does not fully reflect productive wealth.
The variable is kept in this equation to indicate the need for an adequate
measure of productive wealth.

     The linkage of the agricultural sector to earnings and employment in the
overall economy will be considered first.  Exogenous changes in net farm
income and farm labor earnings are included in the total earnings identity,
equation (B-5).  Each dollar increase in farm earnings directly raises per-
         TABLE Bl.   ESTIMATED PARAMETERS FOR INTERINDUSTRY MODEL OF
                           STATE ECONOMIES, 1975
       Variable
Estimated
Parameter
Student
t-value
Equation  (B-l) Dependent Variable - Basic Employment in Logarithms
  Intercept                                        0.0968          9.7864*
  Service employment  (endogenous) in logarithms    0.4279          5.2822*
  Manufacturing value of shipments in logarithms   0.5624          8.6893*

Equation  (B-2) Dependent Variable - Service Employment in Logarithms
  Intercept                                        2.7829
  Basic employment  (endogenous) in logarithms      0.6236
  Total farm input purchases in logarithms
    multiplied by relative importance of
    farming                                        0.1671
  Percentage of population that is metropolitan    0.0101
                 4.7972*
                12.6587*
                 4.8328*
                 5.2911*
Equation (B-6) Dependent Variable - Personal Income in Logarithms
  Intercept                                        1.3286          4.9736*
  Total earnings (endogenous) in logarithms        0.9742         51.1577*
  Wealth in logarithms                             0.0241          1.3672*

Equation (B-8) Dependent Variable - Retail Sales in Logarithms
  Intercept                                        1.6937          4.8624*
  Per capita personal income (endogenous) in
    logarithms                                     0.5510          8.0363*
  Population in logarithms                         0.9667        111.7537*
     *Indicates a statistical significance at a level of 1 percent.

                                      92

-------
sonal income by  $1.30.   Service employment is  directly  related to the level
of farm input purchases.   The variable used to capture  this relationship is
the logarithm of  input  purchases multiplied by the  relative importance of
agriculture in a state's  economy.   Using sample averages, each $1 million in
farm input purchases  directly increased service employment by 106 jobs.

     An examination of  the employment equations indicates that an increase in
basic employment will have a large direct impact on service employment.  Each
one hundred new  jobs  in basic industries will  result in 117 new jobs in ser-
vice industries,  on the margin.  One hundred new jobs in service industries,
however, will require only 23 more workers in  basic industries within the
state.

     Some other  results should also be mentioned.   The  marginal effect of
total earnings is $1.30 in the personal income equation and appears to be
consistent with  the national income accounts because personal income consists
of more than the sum of earnings.   The marginal effect  of personal income on
retail trade, 0.38, is  lower than some estimates of the marginal propensity
to consume, but  this  should be expected because retail  trade is not a com-
plete measure of consumption.

     In addition to explaining economic flows, the  interindustry model can be
used to calculate agricultural production multipliers.   The agricultural pro-
duction multiplier, which shows the change in  personal  income resulting from
a one unit  change in farm sales, was calculated for Georgia and compared with
estimates given  by Schaffer, Laurent, and Sutter.   They reported that each
dollar of final  sales resulted in an increase  of personal income by 71
-------
                                                   APPENDIX C

                                       LAND RESOURCE INVENTORY FOR GEORGIA

                                TABLE  Cl.   SOIL RESOURCE GROUPS (SRGs) OF GEORGIA
VD
-P-
                                  Dominant
                                  Capability
                            %         Class     Hydro-  	Suitability"         Erod.  .      Soil
Name
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Ailey
Americus
Angie
Ardilla
Ashe
Bayboro
Bibb
Bladen
Boswell
Carnegie- 1§
Carnegie-2
Cartecay
Cecil
Chewacla
Conasauga
Cowarts-1
Cowarts-2
Craven
Dotahn
Echaw
Esto-1
Esto-2
Etowah
Fuquay
Faceville
Slope* & Subclass*
5-12
12-25
0-5
0-5
>60
0-5
0-5
0-5
5-12
5-12
12-25
0-5
5-12
0-5
5-12
5-12
12-25
0-5
5-12
0-5
5-12
12-25
5-12
5-12
5-12
4s
7s
2e
2e
7s
5w
5w
5w
4e
4e
4e
5w
2e
4w
6e
4e
6e
3e
2e
5w
4e
6e
3e
3s
2e
" Group**
B
A
C
C
B
D
C
D
D
B
B
C
C
C
C
C
C
C
B
B
B
B
B
B
B
Rower ops
5
5
3
3
6
4
3
5
5
3
3
3
1
4
3
3
4
1
1
5
3
5
1
3
I
Pasture
3
5
3
1
6
3
3
3
3
3
3
1
1
3
3
1
3
1
1
4
3
3
1
1
1
Woodland
3
3
1
1
3
1
1
1
3
1
1
1
1
1
3
1
3
3
1
3
3
3
1
1
1
Factor1"
.20
.17
.32
.24
.24
.17
.20
.10
.43
.28
.28
.28
.32
.32
.43
.24
.24
.37
.24
.10
.32
.32
.37
.20
.37
Classification'1"
Ultisols
Ultisols
Ultisols
Ultisols
Inceptisols
Ultisols
Entisols
Ultisols
Alf isols
Ultisols
Ultisols
Entisols
Ultisols
Inceptisols
Alf isols
Ultisols
Ultisols
Ultisols
Ultisols
Spodosols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
                                                                                             (continued)

-------
                                                TABLE Cl.   (CONTINUED)
VO
Ul
                                 Dominant
                                Capability
                           %       Class    Hydro-
Suitability
           ##
Name
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
Gilead
Goldsboro
Grady
Greenville-1
Greenville-2
Hartsells
Hayesville-1
Hayesville-2
Hector
Helena
Iredell
Istokpoga
Johnson
Kershaw
Kins ton
Lakeland- 1
Lake land- 2
Leef ield
Leon
Louisburg
Marlboro
Mascotte
Meggett
Mountainburg
Nella
Slope* & Subclass
5-12
0-5
0-5
0-5
5-12
5-12
5-12
12-25
25-60
5-12
5-12
0-5
0-5
5-12
0-5
5-12
12-25
0-5
0-5
12-25
5-12
0-5
0-5
25-60
25-60
6e
2w
5w
4e
2e
2e
2e
6e
7e
3e
2e
3w
7w
7s
3w
4s
6s
2w
4w
6e
2e
4w
3w
7s
6s
* Group**
C
B
B
B
B
B
B
B
D
C
D
D
D
A
D
A
A
C
D
B
B
D
D
D
B
Rowcrops
5
1
5
5
1
1
1
4
6
3
1
6
5
5
5
5
5
3
5
6
1
5
4
6
6
Pasture
3
1
5
1
1
1
1
3
3
1
1
6
4
5
2
3
3
3
5
5
1
3
3
5
3
Woodland
3
1
1
1
1
3
1
1
3
3
3
6
1
5
1
3
3
2
5
3
1
3
1
5
3
Factor'
.17
.20
.10
.24
.24
.28
.20
.20
.17
.37
.20
.01
.20
.15
.24
.17
.17
.01
.20
.20
.20
.20
.32
.17
.20
Classification'"
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Inceptisols
Ultisols
Alfisols
Histisols
Ultisols
Entisols
Entisols
Entisols
Entisols
Ultisols
Spodosols
Inceptisols
Ultisols
Spodosols
Alfisols
Ultisols
Ultisols
                                                                                          (continued)

-------
                                             TABLE Cl.  (CONTINUED)
vo
Name
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
Dominant
Capability
% Class Hydro-
Slope* & Subclass # Group**
Norfolk
Ocilla
Oktibeha
Orangeburg-1
Orangeburg-2
Pa co let
Pelham
Saluda
Shack- 1
Shack- 2
Tallapoosa-1
Tallapoosa-2
Tidal Marsh
Tlfton-1
Tifton-2
Townley-1
Town ley- 2
Tr ansy Ivania
Troup-1
Troup-2
Vaucluse-1
Vaucluse-2
Wagram
Urban Area
5-12
0-5
5-12
5-12
12-25
12-25
0-5
>60
5-12
25-60
12-25
25-60
0-5
0-5
5-12
5-12
12-25
0-5
0-5
5-12
5-12
12-25
5-12
0-5
2e
3w
3e
2e
6e
4e
5w
7e
2e
6e
6e
7e
7w
2e
3e
4e
6e
2e
3e
4s
4e
6e
2s
8s
B
C
B
B
B
B
B
C
B
B
C
C
D
B
B
C
C
B
A
A
C
C
A
D
Suitability ##
Rower ops
1
3
3
1
5
4
5
6
1
5
6
6
6
1
1
3
6
1
4
4
5
5
3
6
Pasture
1
2
3
1
1
2
5
6
1
5
5
5
6
1
1
3
3
1
3
3
3
3
1
6
Woodland
1
3
3
1
1
1
1
3
3
3
3
3
6
1
1
3
3
2
3
3
3
3
1
6
Erod.
Fact or "f"
.17
.17
.32
.24
.24
.20
.10
.20
.28
.28
.28
.28
.01
.24
.24
.37
.37
.32
.17
.17
.17
.17
.15
0
Soil
Classification"'"'"
Ultisols
Ultisols
Alfisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Histisols
Ultisols
Ultisols
Ultisols
Ultisols
Inceptisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
Ultisols
                                                                                               (continued)

-------
                                    TABLE Cl.  (CONTINUED)
  % Slope - source is from average slope of soil associations taken from USDA/SCS  soil association
  map.

 "Dominant Capability Class and Subclass - source is from SCS where the class  and  subclass,  which
  best fit the midpoint of the average slope range for each soil association,  was  assigned to that
  complete mapping unit, later to be called a Soil Resource Group.

&&
  Hydro-Group - source is from soil series data provided by SCS.
       Hydrologic soil group  A = low runoff
                              B = moderate runoff
                              C = moderately high runoff
                              D = high runoff

  The different infiltration characteristics of soils are accounted for by classifying soils  into
  four hydrologic soil groups based on the minimum rate of infiltration obtained for bare soil after
  prolonged wetting.

'"'Suitability - source is from published soil survey data.
                              1 = good
               , _             2 = fair-good
           rated for           3 = fa±r
     potential continuous     4 =     _fair
      crop production         5 m poor
                              6 = unsatisfactory
  "^Erodibility Factor - source is from published soil survey data.   Soil-erodibility factor is the
   average soil loss per unit of erosive rainfall (R) under arbitrary selected conditions and depends on
   soil properties.

   Soil Classification - source is from SCS published soil survey data.   All soils belong to one of  10
   major  orders of soils according to an international soil classification  system.

   Some SRG's have the same name but different slopes.

-------
                        TABLE C2. TOTAL HECTARES FOR EACH SOIL RESOURCE GROUP AND LAND RESOURCE
                                                     AREA IN GEORGIA*
00


1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Soil
Resource
Group

Alley
Amerlcus
Angle
Ardilla
Ashe
Bayboro
Bibb
B laden
Boswell
Carnegie-1*
Carnegie-2
Cartecay
Cecil
Chewacla
Conasauga
Cowarts-1
Cowarts-2
Craven
Do than
Echaw
Esto-1
Esto-2
Land' Resource Area
Appalachian Southern
Valley and Coastal
Ridges Mountain Piedmont Sand Hills Plain
(Hectares)
1,238
64,282
15,789
5,167
193,349

31,297 50,009
11,114
19,014
27,284
16,637
58,455 20,249 528,667
2,194,390
15,577 127,258
73,284
84,639 308,589
30,368
15,922
39,489 841,471

44,188
124,093

Atlantic
Coast
Flatwoods Total

1,238
64,282
15,789
5,167
193,349
53,477 53,477
81,306
148,898 160,012
19,014
27,284
16,637
607,371
2,194,390
33,055 175,890
73,284
393,228
30,368
15,922
880,960
27,360 27,360
44,188
124,093
                                                                                         (continued)

-------
TABLE C2. (CONTINUED)



Soil
Resource


23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
Group

Etowah
Fuquay
Faceville
Gilead
Goldsboro
Grady
Greenville-1
Greenville-2
Hartsells
Hayesville-1
Hayesville-2
Hector
Helena
Iredell
Istokpoga
Johnson
Kershaw
Kins ton
Lake land -1
Lakeland- 2
Leefield
Leon
Louisburg
Land Resource Area
Appalachian Southern Atlantic
Valley and Coastal Coast
Ridges Mountain Piedmont Sand Hills Plain Flatwoods
(Hectares)
29,381
115,660
155,475
24,699 130,014
16,193
23,263
20,307
23,639 356,696
20,044
57,965
111,266
19,723
20,394 17,839
41,552
214,798
177,844
60,906 39,139
324,090 24,499
334,144 186,572 137,823
58,576 40,873
66,821 663,980
33,673
115,567



Total

29,381
115,660
155,475
154,713
16,193
23,263
20,307
380,335
20,044
57,965
111,266
19,723
38,233
41,552
214,798
177,844
100,045
348,589
658,539
99,449
730,801
33,673
115,567
(continued)

-------
                                             TABLE C2.  (CONTINUED)
o
o


46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
Soil
Resource
Group

Marlboro
Mascotte
Meggett
Mountainburg
Nella
Norfolk
Ocilla
Oktibbeha
Orangeburg-1
Orangeburg-2
Pacolet
Pelham
Saluda
Shack-1
Shack-2
Tallapoosa-1
Tallapoosa-2
Tidal Marsh
Tifton-1
Tifton-2
Townley-1
Townley-2
Transylvania
Troup-1
Troup-2
Land Resource Area
Appalachian Southern
Valley and Coastal
Ridges Mountain Piedmont Sand Hills Plain
(Hectares)
40,242

13,129 151,623
30,090
139,416
425,621

21,663
7,181 426,841
82,251
1,286,755
9,850
728 107,121
107,905
179,254
157,222
138,966


1,184,636
32,180
112,144
3,731

62,040

Atlantic
Coast
Flatwoods Total

40,242
616,996 616,996
19,696 184,448
30,090
139,416
425,621
73,389 73,389
21,663
434,022
82,251
1,286,755
42,351 52,201
107,849
107,905
179,254
157,222
138,966
207,091 207,091
153,517 153,517
1,184,636
32,180
112,144
3,731
39,455 39,455
62,040
                                                                                       (Continued)

-------
                                         TABLE C2.  (CONTINUED)





71.
72.
73.
74.

Soil Appalachian
Resource Valley and
Group Ridges

Vacluse-1
Vacluse-2
Wagram
Urban area
Total 752,791
Land Resource Area
Southern
Coastal
Mountain Piedmont Sand Hills Plain
(Hectares)
3,560
112,573
175,345
19,594
652,370 4,394,231 766,488 5,779,259

Atlantic
Coast
Flatwoods Total

3,560
112,573
175,345
19,594
2,707,041 15,052,180
*Source:   U. S. Department of Agriculture,  Soil Conservation Service.
          Georgia.  Athens, Georgia,  1977.

#Some Soil Resource Groups have the same name but different  slopes.
Soil Association Map of

-------
     TABLE C3.  CULTIVATABLE,  MARGINAL, WETLANDS AND SUBMARGINAL LANDS IN ROW CROPS BY LAND RESOURCE AREA*
o
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total

Distribution
Cultivatable
(1,11,111)


53,215
10,150
111,217
39,530

1,036,867

149,536
1,400,515

93.7
Marginal
(IV)


3,132
1,274
15,351
3,860

37,445

9,232
70,294

4.7
Distribution by
Wetlands Submarginal Land Resource
(V) (VI, VII, VIII) Total Area
(Hectares)

0
0
181
156

5,553

1,850
7,740
(Percent)
0.5


1,172
1,443
5,655
787

6,902

459
16,418

1.1


57,519
12,867
132,404
44,333

1,086,767

161,077
1,494,967

100
(Percent)

3.8
0.9
8.8
3.0

72.7

10.8
100


     *Source:   U.S.  Department  of Agriculture,  Soil Conservation Service.  Georgia Conservation Needs
      Inventory,  1967.   Washington, D.C., April, 1970.

-------
    TABLE 04. CULTIVATABLE, MARGINAL, WETLANDS AND  SUBMARGINAL LANDS IN FOREST BY LAND RESOURCE AREA*
o
U)
Land Resource
Area

Appalachian Valley
and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal
Plain
Atlantic Coast
Flatwoods
Total

Distribution
Cultivatable
(I, II, III)


114,136
35,272
1,080,431
191,261

1,600,516

620,179
3,641,795

37.9
Marginal
(IV)


68,555
48,421
815,374
110,692

346,396

84,465
1,473,903

15.4
Wetlands
(V)
(Hectares)

0
234
12,996
44,331

895,476

980,678
1,933,715
(Percent)
20.1
Percent in Each
Submarginal Land Resource
(VI, VII, VIII) Total Area


281,453
296,744
1,035,827
142,654

567,246

228,361
2,552,285

26.6


464,144
380,671
2,944,628
488,938

3,409,634

1,913,683
9,601,698

100
(Percent)

4.8
4.0
30.7
5.1

35.5

19.9
100


     *Source:  U.S. Department of Agriculture,  Soil Conservation Service.  Georgia Conservation Needs

      Inventory, 1967.  Washington, D.C., April,  1970.

-------
     TABLE C5. CULTIVATABLE, MARGINAL, WETLANDS AND SUBMARGINAL LANDS IN OTHER USES* BY LAND RESOURCE AREA//
o
Land Resource Cultivatable
Area (I, II, III)

Appalachian Valley and
Ridges 96,738
Mountain 37,968
Piedmont 657,080
Sand Hills 88,046
Southern Coastal
Plain 959,323
Atlantic Coast
Flatwoods 127,425
Total 1,966,580

Distribution 72.6
Percent in Each
Marginal Wetlands Submarginal Land Resource
(IV) (V) (VI, VII, VIII) Total Area
(Hectares)
35,142 0 30,866 162,746
12,330 0 16,948 67,246
189,158 2,186 108,246 956,670
16,067 2,216 5,693 112,022

87,302 27,169 30,557 1,104,351

9,608 21,654 147,813 306,500
349,607 53,225 340,123 2,709,535
(Percent)
12.9 2.0 12.5 100
(Percent)
6.0
2.5
35.3
4.1

40.8

11.3
100


     *0ther Uses as used herein include land for conservation use only, pasture and rangeland, orchards,
      close grown crops and idle land.

     //Source:  U.S. Department of Agriculture, Soil Conservation Service.  Georgia Conservation Needs
      Inventory, 1967.  Washington, D.C., April, 1970.

-------
o
Ln
                                                    APPENDIX D



                                          RUNOFF AND SEDIMENT YIELD DATA





                      TABLE Dl.   RUNOFF CURVE NUMBERS (CN) FOR SELECTED AGRICULTURAL USES*
Land Use Description
Cultivated land without conservation
treatment
Cultivated land with conservation
treatment
Pasture, range land, meadow
Wood or forest land; moderate
cover
Hydrologic
A B
72 81
62 71
54 70
35 61
Soil Group
C
88
78
80
74

D
91
81
84
80
        *Source:  Georgia State Soil and Water Conservation Committee.   Manual for Erosion and Sediment

         Control in Georgia, Athens, Georgia, 1977.

-------
TABLE D2.  RAINFALL EROSION INDEX FACTOR "R" VALUES*
County
Appling
Atkinson
Bacon
Baker
Baldwin
Banks
Barrow
Bar tow
Ben Hill
Berrien
Bibb
Bleckley
Brant ley
Brooks
Bryan
Bulloch
Burke
Butts
Calhoun
Camden
Candler
Carroll
Catoosa
Char 1 ton
Chatham
Chattahoochee
Chattooga
Cherokee
Clarke
R
350
350
350
350
250
300
300
300
350
350
300
300
350
350
350
300
250
300
350
350
300
300
250
350
350
350
300
300
250
County
Clay
Clayton
Clinch
Cobb
Coffee
Colquitt
Columbia
Cook
Coweta
Crawford
Crisp
Dade
Daws on
Decatur
Dekalb
Dodge
Dooly
Dougherty
Douglas
Early
Echols
Eff ingham
Elbert
Emanuel
Evans
Fannin
Fayette
Floyd
Forsyth
R
350
300
350
300
350
350
250
350
300
300
350
250
300
350
300
300
300
350
300
350.
350
350
250
300
350
250
300
300
300
County
Franklin
Fulton
Gilmer
Glascock
Glynn
Gordon
Grady
Greene
Gwinnett
Habersham
Hall
Hancock
Kara Is on
Harris
Hart
Heard
Henry
Houston
Irwin
Jackson
Jasper
Jeff Davis
Jefferson
Jenkins
Johnson
Jones
Lamar
Lanier
Laurens
R
300
300
300
250
350
300
350
250
300
350
300
250
300
350
300
350
300
300
350
300
300
350
250
300
300
300
300
350
300
                                                   (continued)
                       106

-------
TABLE D2.  (CONTINUED)
County
Lee
Liberty
Lincoln
Long
Lowndes
Lump kin
McDuffie
Mclntosh
Macon
Madison
Marion
Meriwether
Miller
Mitchell
Monroe
Montgomery
Morgan
Murray
Muscogee
Newton
Oconee
Oglethorpe
Paulding
Peach
Pickens
Pierce
Pike
Polk
Pulaski
R
350
350
250
350
350
300
250
350
300
250
350
300
350
350
300
300
250
250
350
300
250
250
300
300
300
350
300
300
300
County
Putnam
Quitman
Rabun
Randolph
Richmond
Rockdale
Schley
Screven
Seminole
Spalding
Stephens
Stewart
Sumter
Talbot
Taliaferro
Tattnall
Taylor
Telfair
Terrell
Thomas
Tift
Toombs
Towns
Treutlen
Troup
Turner
Twiggs
Union
Up son
R
250
350
350
350
250
300
350
300
350
300
300
350
350
350
250
350
300
350
350
350
350
350
300
300
350
350
300
300
30.0
County
Walker
Walton
Ware
Warren
Washington
Wayne
Webster
Wheeler
White
Whitfield
Wilcox
Wilkes
Wilkinson
Worth















R
250
300
350
250
250
350
350
300
300
250
350
250
300
350
















-------
        TABLE D3.  PLANT COVER OR CROPPING MANAGEMENT FACTOR  (C)*
 Land Use Groups
       Examples
Range of "C" values
Permanent vegetation
Established meadows
Small grains
Large-seeded legumes
Row crops
Fallow
Protected woodland
Prairie
Permanent pasture
Sodded orchard
Permanent meadow

Alfalfa
Clover
Fescue

Rye
Wheat
Barley
Oats

Soybeans
Cowpeas
Peanuts
Field peas

Cotton
Potatoes
Tobacco
Vegetables
Corn
Sorghum

Summer fallow
Period between plowing and
  growth of crop
    0.0001-0.45
    0.004-0.3
    0.07-0.5
    0.01-0.65
    0.1-0.70
    1.0
*Source:  Georgia State Soil and Water Conservation Committee.  Manual for
 Erosion and Sediment Control in Georgia, Athens, Georgia, 1977.
                                     108

-------
                TABLE D4.   EROSION-CONTROL PRACTICE FACTOR*
Land Slope
%
2.0-7
8.0-12
13.0-18
19.0-24
P Values
Contouring
0.50
0.60
0.80
0.90

Terracing*
0.10
0.12
0.16
0.18
*Source:  Georgia State Soil and Water Conservation Committee.  Manual for
 Erosion and Sediment Control in Georgia, Athens, Georgia, 1977.

#For prediction of contribution to off-field sediment load.
                                      109

-------
        TABLE D3.  PLANT COVER OR CROPPING MANAGEMENT FACTOR  (C)*
 Land Use Groups
       Examples
Range of "C" values
Permanent vegetation
Established meadows
Small grains
Large-seeded legumes
Row crops
Fallow
Protected woodland
Prairie
Permanent pasture
Sodded orchard
Permanent meadow

Alfalfa
Clover
Fescue

Rye
Wheat
Barley
Oats

Soybeans
Cowpeas
Peanuts
Field peas

Cotton
Potatoes
Tobacco
Vegetables
Corn
Sorghum

Summer fallow
Period between plowing and
  growth of crop
    0.0001-0.45
    0.004-0.3
    0.07-0.5
    0.01-0.65
    0.1-0.70
    1.0
*Source:  Georgia State Soil and Water Conservation Committee.  Manual for
 Erosion and Sediment Control in Georgia, Athens, Georgia, 1977.
                                     108

-------
                TABLE D4.   EROSION-CONTROL PRACTICE FACTOR*
Land Slope
%
2.0-7
8.0-12
13.0-18
19.0-24
P Values
Contouring
0.50
0.60
0.80
0.90

Terracing''
0.10
0.12
0.16
0.18
*Source:  Georgia State  Soil and Water  Conservation Committee.  Manual for
 Erosion and Sediment Control in Georgia, Athens, Georgia, 1977.

#For prediction of contribution to off-field sediment load.
                                     109

-------
                     TABLE D5.  ESTIMATED SEDIMENT DELIVERY RATIO AS RELATED TO
                              AVERAGE DRAINAGE AREA FOR PERENNIAL STREAM
Land Resource Area
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Average Drainage
Area (Square Hectares)*
0.97
0.12
0.58
1.93
1.55
2.70
Sediment
Delivery Ratio*
0.25
0.36
0.27
0.12
0.13
0.10
*Source:  Basic watershed map established for Georgia in 1959 by SCS, Athens, Georgia.

^Source:  J. W. Roehl.  Sediment Source Areas, Delivery Ratios, and Influencing Morphological
 Factors.  Commission on Land Erosion.  Publication No. 59.  International Association of
 Scientific Hydrology, 1962.

-------
   20.0
  Slope Length, Meters

•20        40   60    100
                                                             200     400   600
                                                                 i _!^'>' l  i
                              60     TOO      200

                                  Slope Length, Feet
                      400   600   1000
2000
                      Figure Dl.  Slope-effect chart.*

*The dashed  lines  represent estimates for slope dimensions  beyond  the  range
 of lengths  and  steepnesses for which data are available.
                                     Ill

-------
                         APPENDIX E

                 ENVIRONMENTAL IMPACT ON ALL
                  NEW CROPLAND IN GEORGIA
TABLE El.  ESTIMATED RUNOFF FROM NEW CROPLAND IN GEORGIA, 1976
Land Resource
Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total State
All New
Runoff
(cm)

6.73
9,98
9.47
10.08
8.86
11.38
9.07

9.04
9.60
7.62
5.44
8.61
12.24
9.63
9.22
Cropland
Increased
Runoff
(.cm)

1,55
1.70
0.71
2.03
2.85
3.91
2.77

0.64
0.89
1.30
2.61
2.51
3.35
2,69
2.77
Soil
Runoff
(cm)

6.73
9.91
9. '43
10.74
9.04
11.38
9.07

8.99
^
7.80
7.24
8.66
10.49
8.61
9.02
Bank Land
Increased
Runoff
(cm)

1.14
1.50
0.36
1.02
1.45
2.41
1.17

0.33
-
1.04
1.70
1.47
2.16
1.37
1.19

-------
TABLE E2.  ESTIMATED INCREASED SEDIMENT YIELD PROM NEW CROPLAND IN GEORGIA,  1976
Land Resource
Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total State
All New
Total
(1,000 MT)

123.86
4.05
431.02
26.88
1,769.37
54.04
2,409.22

16.16
2,11
96.96
16.93
193,52
8,68
334,36
2,743.58
Cropland
Average
per Hectare
(MT)

27.60
23.03
24.64
7.48
8.76
1.93
9.43

31.20
60.44
31.56
8.11
10.35
0.67
8.92
9.36
Soil Bank Land
Total
(1,000 MT)

68.88
2.35
225.34
8.54
218.84
1.74
525.69

8.82
0.00
30.11
2.11
23.87
0.63
65.54
591.23
Average
per Hectare
(MT)

21.10
22.80
25.02
7.59
9.77
1,64
14.22

29.14
,_
30.31
9.77
13.08
1.28
17.14
14.49

-------
TABLE E3.  ESTIMATED SEDIMENT YIELD FROM NEW CROPLAND IN GEORGIA, 1976
Land Resource
Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total State
All New
Total
(1,000 MT)

128.29
4.18
446.95
28.21
1,865.45
56.95
2,530.03

16.71
2,17
99.91
17,59
203,29
9,21
348.88
2,878,91
Cropland
Average
per Hectare
(MT)

28.58
23.74
25.56
7.86
9.25
2.04
9.90

32.26
62,07
32,52
8.42
10.86
0.69
9.30
9.81
Soil
Total
(1,000 MT)

70.84
2.41
231.78
8.79
225.06
1.78
540.66

9.05
0.00
30.86
2.16
24.50
0.64
67.21
607.87
Bank Land
Average
per Hectare
(MT)

21.71
23.45
25.74
7.82
10.06
1.68
14.63

29.95
-
31.07
9.99
13.42
1.32
17.58
14.90

-------
TABLE E4.   ESTIMATED NITROGEN LOSS FROM NEW CROPLAND IN GEORGIA, 1976
Land Resource
Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods.
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total State
All
Total
(1,000 Kg)

48.76
3.85
205.54
22.50
1,765,76
128.19
2,174,30

4.62
1.82
44.81
8.94
156.09
52,14
268.42
2,442.72
New Cropland
Average
per Hectare
(Kg)

10.88
21.86
11.76
6.18
8.76
4.58
8.51

8.94
52.01
14.60
4.28
8.36
3.98
7.16
8.34
Soil
Total
(1,000 Kg)

23.14
2.29
87.97
5.93
213.50
4.16
336.99

2.00
0.00
9.98
1.46
21.86
1.88
37.18
374.17
Bank Land
Average
per Hectare
(kg)

7.10
22.24
9.78
5.28
9.54
3.94
9.13

6.62
_
10.05
6.74
12.00
3.85
9.73
9.18

-------
TABLE E5.   ESTIMATED PHOSPHORUS LOSS FROM NEW CROPLAND IN GEORGIA, 1976
Land Resource
Area

Land Suitable for Cultivation
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Cultivatable
Marginal and Submarginal Land
Appalachian Valley and Ridges
Mountain
Piedmont
Sand Hills
Southern Coastal Plain
Atlantic Coast Flatwoods
Total Marginal and Submarginal
Total State
All New
Total
(1,000 Kg)

105.18
4.14
375,63
25.20
1,783.39
56.07
2,349.61

13.05
2.15
84.13
14.79
186,96
8.66
309.74
2,659.35
Cropland
Average
per Hectare
(Kg)

23.45
23.58
21.49
7.02
8.85
2,00
9.21

25.21
61.49
27.42
7.09
10.00
0.66
8,25
9.08
Soil
Total
(1,000 Kg)

56.48
2.42
188.51
7.53
216.19
1.72
472.85

6.89
0.00
24,48
1.93
23,61
0.60
57.51
530.36
Bank Land
Average
per Hectare
(Kg)

17.32
23.51
20.94
6.70
9.67
1.62
12.81

22.81
-
24.67
8.92
12.95
1.23
15.05
13.01

-------
1. REPORT NO.
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 EPA-600/5-80-OQ2
                              2.
                                                            3. RECIPIENT'S ACCESSION-NO.
 Environmental  and Economic Impact of Agricultural Land
 Use Conversion:   An Evaluation Methodology
              5. REPORT DATE
               January 1980 issuing date
              6. PERFORMING ORGANIZATION CODE
 F.C. White, J.E.  Hairston, W.N.  Musser, H.F. Perkins
 and J.F. Reed                                         '
                                                            8. PERFORMING ORGANIZATION REPORT NO.
                             XND ADDRESS
 College of Agriculture
 University of  Georgia
 Athens, Georgia  30602
              10. PROGRAM ELEMENT NO.
                A34B1B
              11. CONTRACT/GRANT NO.
                R804510-01
 12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental  Research Laboratory—Athens
 Office of Research and Development
 U.S. Environmental Protection Agency
 Athens, Georgia  30605
GA
13. TYPE OF REPORT AND PERIOD COVERED
  Final,  7/76-8/78
              14. SPONSORING AGENCY CODE
                EPA/600/01
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT                                                  ~—~~	—~~"	        	
       The development and application of a methodology for evaluating the environmental
 and economic  impacts of placing marginal,  submarginal, and Soil Bank land in crop pro-
 duction is described.  Environmental  impacts were measured by quantifying the increased
 environmental loadings of sediment, nitrogen, and phosphorus and by estimating the po-
 tential environmental impact of pesticides.  Economic consequences considered included
 changes in net farm income and aggregate impact on the state's economy.  Although no
 attempt was made to place a dollar value on environmental consequences, these impacts
 can be weighed against the economic impacts in a tradeoff fashion as a measure of socia
 costs and benefits resulting from increased production on new cropland.

       The developed methodology was applied to the state of Georgia for the period
 1973-1976.  The pollutants generated  annually from each hectare of converted crop pro-
 duction land  during the period were estimated to be 11.9 metric tons of sediment, 8.1
 kilograms of  nitrogen loss (excluding leaching), and 10.5 kilograms of phosphorus loss.
 Net farm income increased an estimated $61 per hectare annually.

       Methodology used in this study  could be used in similar studies in other areas
 of the United States.  Much of the required data is already available in other areas.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                           c. COS AT I Field/Group
 Water pollution
 Agricultural economics
 Planning
 Analysis
                             02B
                             68D
                             91A
 8. DISTRIBUTION STATEMENT

 RELEASE TO PUBLIC
  UNCLASSIFIED
                 127
20. SECURITY CLASS (Thispage)
  UNCLASSIFIED
                           22. PRICE
EPA Form 2220-1 (9-73)
                                            117
                                                                  « US GOVERNMENT PRINTING OFf ICE 1980-657-146/5551

-------