&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 on
the average. As expected, the marginal effect of farm sales is somewhat
lower than this average estimate. The agricultural production multiplier,
which for this model may be interpreted as showing both the direct and
indirect change in personal income resulting from a dollar change in farm
sales, was estimated to be 51<: on the margin for Georgia.
93
-------
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
------- |