• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Is agricultural sustainability...
 A systems framework for characterizing...
 An object-oriented representation...
 Crop simulation for characterizing...
 Determinants of sustainability...
 Summary and conclusions
 Appendix A: Object-oriented programming...
 Appendix B: A minimum data set...
 Appendix C: Farming system simulator...
 Appendix D: Input files used for...
 Bibliography
 Biographical sketch






Title: A systems approach to characterizing farm sustainability
CITATION PAGE IMAGE ZOOMABLE PAGE TEXT
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00056221/00001
 Material Information
Title: A systems approach to characterizing farm sustainability
Physical Description: xix, 265 leaves : ill. ; 29 cm.
Language: English
Creator: Hansen, James William, 1960-
Publication Date: 1996
 Subjects
Subject: Agricultural systems   ( lcsh )
Sustainable agriculture   ( lcsh )
Agricultural and Biological Engineering thesis, Ph. D
Dissertations, Academic -- Agricultural and Biological Engineering -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1996.
Bibliography: Includes bibliographical references (leaves 249-264).
Statement of Responsibility: by James William Hansen.
General Note: Typescript.
General Note: Vita.
Funding: Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
 Record Information
Bibliographic ID: UF00056221
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 002097373
oclc - 35010234
notis - AKT6148

Table of Contents
    Title Page
        Page i
        Page ii
    Acknowledgement
        Page iii
    Table of Contents
        Page iv
        Page v
        Page vi
        Page vii
    List of Tables
        Page viii
        Page ix
        Page x
        Page xi
    List of Figures
        Page xii
        Page xiii
        Page xiv
        Page xv
        Page xvi
        Page xvii
    Abstract
        Page xviii
        Page xix
    Introduction
        Page 1
        Page 2
        Page 3
    Is agricultural sustainability a useful concept?
        Page 4
        Introduction
            Page 4
            Page 5
        Sustainability as an approach to agriculture
            Page 6
            Sustainability as an alternative ideology
                Page 7
                Page 8
                Page 9
                Page 10
            Sustainability as a set of strategies
                Page 11
                Page 12
                Page 13
            Discussion
                Page 14
                Page 15
                Page 16
        Sustainability as a property of agriculture
            Page 17
            Sustainability as a an ability to satisfy goals
                Page 17
            Sustainability as an ability to continue
                Page 18
        Approaches to characterizing sustainability
            Page 19
            Adherence to prescribed approaches
                Page 19
            Multiple qualitative indicators
                Page 20
            Integrated, qualitative indicators
                Page 21
                Page 22
            Time trends
                Page 23
                Page 24
            Resilience
                Page 25
            System simulation
                Page 26
        Elements of a useful approach for characterizing sustainability
            Page 27
            Page 28
            Page 29
            Page 30
            Page 31
        Conclusions
            Page 32
    A systems framework for characterizing farm sustainability
        Page 33
        Introduction
            Page 33
        Defining sustainability
            Page 34
            Page 35
        Quantifying sustainability
            Page 36
            Failure as violation of state thresholds
                Page 36
                Page 37
                Page 38
                Page 39
                Page 40
            Sustainability hazard
                Page 41
        Simulating sustainability
            Page 42
            Page 43
            An example: Sustainability of a simple time-series model
                Page 44
                Page 45
        Diagnosing constraints to sustainability
            Page 46
            Page 47
            Sensitivity analysis
                Page 48
            Significance tests
                Page 49
                Page 50
                Page 51
        Sustainability applied to farming systems
            Page 52
            Selecting a time frame
                Page 53
            Assumptions about inputs
                Page 54
            Farm failure criteria
                Page 55
            Determinants of farm sustainability
                Page 56
                Page 57
        An example: Sustainability of a coastal Texas rice farm
            Page 58
            Methods
                Page 59
            Results
                Page 60
                Page 61
        Discussion
            Page 62
            Page 63
            Page 64
    An object-oriented representation of a farming system
        Page 65
        Introduction
            Page 65
            Page 66
        Overview of the farming system simulator
            Page 67
            Page 68
            Page 69
            Page 70
            Page 71
        Inputs
            Page 72
            Scenario file
                Page 73
                Page 74
                Page 75
                Page 76
                Page 77
                Page 78
            Price file
                Page 79
            Crop minimum data set
                Page 80
        Processes
            Page 80
            Overview
                Page 81
            Random number sequences
                Page 82
            Price generation
                Page 82
                Page 83
                Page 84
            Crop and ecosystem processes
                Page 85
            Event handling
                Page 86
                Page 87
            Resource accounting
                Page 88
                Page 89
                Page 90
                Page 91
                Page 92
                Page 93
                Page 94
            Illustration: A fertilizer application operation
                Page 95
                Page 96
                Page 97
            Household consumption
                Page 98
        Outputs
            Page 99
            Resource status
                Page 99
                Page 100
            Sustainability
                Page 101
        Discussion
            Page 101
            Page 102
            Page 103
    Crop simulation for characterizing sustainability of a Colombian hillside farm
        Page 104
        Introduction
            Page 104
            Page 105
        A Colombian hillside environment
            Page 106
            Page 107
            Page 108
            Page 109
            Page 110
            Page 111
            Page 112
            Page 113
            Page 114
            Page 115
        Crop simulation
            Page 116
            Page 117
            Page 118
            Page 119
        Approach
            Page 120
            Weather data
                Page 120
                Page 121
                Page 122
                Page 123
                Page 124
            Soil data
                Page 125
            Simulation conditions
                Page 126
            Development and yield
                Page 127
            Response to environmental factors
                Page 128
        Results and discussion
            Page 129
            Crop development yield
                Page 129
                Page 130
                Page 131
                Page 132
                Page 133
                Page 134
                Page 135
            Weather variability
                Page 136
                Page 137
                Page 138
            Response to nitrogen dynamics
                Page 139
                Page 140
                Page 141
                Page 142
                Page 143
                Page 144
                Page 145
                Page 146
                Page 147
                Page 148
                Page 149
                Page 150
                Page 151
            Response to soil erosion
                Page 152
                Page 153
                Page 154
                Page 155
                Page 156
                Page 157
        Linking models
            Page 158
            Issues in linking crop and whole-farm models
                Page 158
                Page 159
                Page 160
            Issues linking crop and erosion models
                Page 161
                Page 162
                Page 163
        Conclusions
            Page 164
            Page 165
            Page 166
    Determinants of sustainability of a Colombian hillside farm
        Page 167
        Introduction
            Page 167
            Page 168
            Page 169
        Approach
            Page 170
            A Colombian hillside farm
                Page 170
            Sources of information
                Page 171
                Page 172
                Page 173
                Page 174
                Page 175
            Assumptions
                Page 176
                Page 177
                Page 178
                Page 179
            Scenarios
                Page 180
                Page 181
                Page 182
                Page 183
                Page 184
            Simulation and analysis
                Page 185
        Results
            Page 186
            Base scenario
                Page 186
                Page 187
                Page 188
                Page 189
            Cropping systems
                Page 190
                Page 191
            Soil management
                Page 192
                Page 193
            Costs and prices
                Page 194
                Page 195
            Resources
                Page 196
                Page 197
            Sources of risk
                Page 198
                Page 199
                Page 200
                Page 201
            Constraints to sustainability
                Page 202
        Discussion
            Page 202
            Practical implications
                Page 203
            Limitations
                Page 204
                Page 205
    Summary and conclusions
        Page 206
        Page 207
        Page 208
        Page 209
    Appendix A: Object-oriented programming concepts
        Page 210
        Page 211
        Page 212
    Appendix B: A minimum data set for simulating farm sustainability
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
    Appendix C: Farming system simulator user's guide
        Page 235
        Page 236
        Page 237
    Appendix D: Input files used for farm simulations
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
    Bibliography
        Page 249
        Page 250
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
    Biographical sketch
        Page 265
        Page 266
        Page 267
Full Text










A SYSTEMS APPROACH TO CHARACTERIZING FARM SUSTAINABILITY


By
JAMES WILLIAM HANSEN

















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1996


























Copyright 1996

by

James William Hansen














ACKNOWLEDGMENTS


I would like first to express my gratitude to my advisor and mentor, Dr. Jim Jones.

He took a risk by giving me the freedom to pursue an idea, then guided and encouraged

me through the process of making it reality. His competence, integrity and compassion

serve as an example to many. Dr. Robert Caldwell first suggested to me the idea that

became the basis for this study. Dr. Phillip Thornton helped me numerous times to clarify

and express my ideas. He and Drs. Peart, Boggess and Hildebrand provided

encouragement and wise guidance for this research. I owe a debt of gratitude to Dr. Ron

Knapp for his kindness and hospitality during my visits to CIAT, and for many long hours

of hunting down information for this research. His enthusiasm helped me to maintain

mine. Thanks goes also to his staff, especially Jorge and Jorge. Jos6 Domingo deserves

honor and thanks. His successful effort to extract a livelihood for his family and higher

education for his children from a few hectares of hilly land provided the basis for much of

this study.

I am grateful to my wife, Merlie, who endured with me and remained my best

friend through the process. In her love I found strength as I worked and a refuge as I

rested. Thanks also to friends at North Central Baptist Church and others who supported

me in prayer. Ultimate thanks goes to the Lord Jesus who gave me strength to complete

this study and compassion that gives it purpose.















TABLE OF CONTENTS


ACKNOWLEDGMENTS .........


. .. . . . . . . iii


LIST OF TABLES ....................... ........... ......

LIST OF FIGURES ......... ........ .................... ...

ABSTRACT ............................. ... ...... ........

CHAPTERS

1 INTRODUCTION .......................................


2 IS AGRICULTURAL SUSTAINABILITY A USEFUL CONCEPT?

Introduction ...........................................
Sustainability as an Approach to Agriculture .................
Sustainability as an Alternative Ideology ....................
Sustainability as a Set of Strategies ........................
D discussion ........................................
Sustainability as a Property of Agriculture .....................
Sustainability as an Ability to Satisfy Goals ..................
Sustainability as an Ability to Continue .....................
Approaches to Characterizing Sustainability ....................
Adherence to Prescribed Approaches ......................
M multiple Qualitative Indicators ...........................
Integrated, Quantitative Indicators ...................
Time Trends ....................................
R esilience ............... ...........................
System Sim ulation ....................................
Elements of a Useful Approach for Characterizing Sustainability ....
C conclusions .........................................


. .. .. 4


. viii

.xii

.xviii


. . 1











3 A SYSTEMS FRAMEWORK FOR CHARACTERIZING FARM
SUSTAINABILITY ....................................... 33

Introduction ................................................. 33
Defining Sustainability ......................................... 34
Quantifying Sustainability .......................... ............. 36
Failure as Violation of State Thresholds ........................ 36
Sustainability Hazard ............................ ............ 41
Simulating Sustainability .............................. ............ 42
An Example: Sustainability of a Simple Time-series Model ............. 44
Diagnosing Constraints to Sustainability ...................... ........ 46
Sensitivity Analysis ......................................... 48
Significance Tests ..................... ....... ................ 49
Sustainability Applied to Farming Systems ........ ............ .52
Selecting a Time Frame .................. .................... 53
Assumptions about Inputs .................................... 54
Farm Failure Criteria ......................................... 55
Determinants of Farm Sustainability ............................. 56
An Example: Sustainability of a Coastal Texas Rice Farm ................. 58
Methods ................................................. 59
Results .................................................... 60
Discussion ........................ ......................... 62

4 AN OBJECT-ORIENTED REPRESENTATION OF A FARMING
SYSTEM ................................................ 65

Introduction .................................................. 65
Overview of the Farming System Simulator ........................... 67
Inputs .............. ............... ....................... 72
Scenario File ............................................. 73
Price File ................ ..... ..... .................... 79
Crop Minimum Data Set .................................... 80
Processes ............ ............................ .. ... ... .......... 80
Overview .............................................. 81
Random Number Sequences .. .................................. 82
Price Generation .......................................... 82
Crop and Ecosystem Processes ............................. ... 85
Event Handling ................................ ............ 86
Resource Accounting ....................................... 88
Illustration: A Fertilizer Application Operation .................. 95
Household Consumption ..................................... 98









Outputs ................. ...... .. .... ............. 99
Resource Status ...................................... ....... 99
Sustainability ............ .... ................. ........ 101
Discussion ............................................... 101


5 CROP SIMULATION FOR CHARACTERIZING SUSTAINABILITY
OF A COLOMBIAN HILLSIDE FARM ....................... 104

Introduction ................. ........... ... .................. 104
A Colombian Hillside Environment ................................ 106
Crop Simulation ................... ................. ......... 116
Approach .......... .............. ....................... 120
Weather Data ........................................... 120
Soil D ata .................................. ...... ...... 125
Simulation Conditions .............. ............. ........ 126
Development and Yield .................................. 127
Response to Environmental Factors ...................... .. ... 128
Results and Discussion ................... .. .... ..... .... . 129
Crop Development and Yield .................................. 129
Weather Variability ................. ....................... 136
Response to Nitrogen Dynamics ................................ 139
Response to Soil Erosion ................... .............. 152
Linking Models ............................................. 158
Issues in Linking Crop and Whole-farm Models .......... ... ... 158
Issues in Linking Crop and Erosion Models ....................... 161
Conclusions ............................................... 164

6 DETERMINANTS OF SUSTAINABILITY OF A COLOMBIAN
HILLSIDE FARM .......................................... 167

Introduction ....................... ........................ 167
Approach .................................................. 170
A Colombian Hillside Farm ................................... 170
Sources of Information ..................................... 171
Assumptions ............................................ 176
Scenarios .................................... ........... 180
Simulation and Analysis ...................................... 185
Results .................................................... 186
Base Scenario ................... ......................... 186
Cropping Systems ...................................... 190
Soil Management .......................................... 192
Costs and Prices ................... ....................... 194










R sources .................................
Sources of Risk ...............................
Constraints to Sustainability ......................
D iscu ssion ......................................
Practical Im plications ...........................
L im stations .................................

7 SUMMARY AND CONCLUSIONS ..................

APPENDICES

A OBJECT-ORIENTED PROGRAMMING CONCEPTS ....

B A MINIMUM DATA SET FOR SIMULATING FARM
SUSTAINABILITY ............................

C FARMING SYSTEM SIMULATOR USER'S GUIDE ....

D INPUT FILES USED FOR FARM SIMULATIONS ......

BIBLIOGRAPHY ......................................

BIOGRAPHICAL SKETCH ..............................


196
198
202
202
203
204


. .......... 2 13


235

238














LIST OF TABLES


Table page

2-1 Interpretations of agricultural sustainability ............................ 5

2-2 Contrasting approaches of conventional and sustainable agriculture as
characterized by Hill and MacRae ................................... 9

2-3 Strategies frequently associated with sustainability ...................... 11

2-4 Contingency table for inferring sustainability based on trends of system inputs
and outputs ................................................... 23

2-5 Elements of a useful approach to characterizing sustainability of agricultural
systems .................................................... 28

2-6 Approaches to characterizing agricultural sustainability .................. '31

3-1 Definitions of symbols used ....................................... 37

3-2 Sensitivity of simulated sustainability to system properties ................ 46

3-3 Frequency table for tests of difference between simulated sustainabilities,
independent observations ......................................... 50

3-4 Frequency table for testing differences in sustainabilities: paired observations .. 51

3-5 Relative sensitivity of simulated five-year sustainability of a Texas rice farm
to continuous factors ............................................ 60

3-6 Absolute sensitivity of simulated five-year sustainability of a Texas rice farm
to discrete factors .............................................. 61

4-1 Sections in the farm scenario file ................................... 74










4-2 Description of resource classes ........................ ............. 77

4-3 Format of the operations output file ............................... 87

5-1 Area in each slope class, Domingo farm, Cauca, Colombia ............... 106

5-2 Properties of soil layers, site near on-farm trials, Domingo farm, Cauca,
Colombia ..................................... ............ 112

5-3 Spatially interpolated WGEN coefficients for the Domingo farm, Cauca,
Colombia ........................... .... ................... 125

5-4 Planting information for crop simulation studies ....................... 126

5-5 Observed and predicted bean yields, Domingo and Trujillo farms, Cauca,
Colombia ................................................. 130

5-6 Observed and predicted timing of phenological events for bean and maize,
Domingo farm, Cauca, Colombia, planted March 30, 1994 ............... 130

5-7 Treatment description and observed and predicted yields for the October 14,
1993 planting of the bean fertility trial, Domingo farm, Cauca, Colombia .... 132

5-8 Treatment description and observed and predicted yields for the March 30,
1994 planting of the bean fertility trial, Domingo farm, Cauca, Colombia .... 132

5-9 Observed and predicted maize yields, Domingo and Trujillo farms, Cauca,
Colombia ................................................ 134

5-10 Observed and predicted cassava yields, Domingo and Trujillo farms, Cauca,
C olom bia ........................................... ......... 135

5-11 Mean, standard deviation, skewness and Kolmogorov-Smirnov test statistic
for distributions of crop yield and maturity time simulated with observed and
simulated weather from La Florida, Cauca, Colombia .................. 137

5-12 Mean, standard deviation, skewness, and Kolmogorov-Smirnov test statistic for
distributions of observed and simulated monthly and annual rainfall totals, La
Florida, Popayan, Colombia ...................................... 139

6-1 Hypotheses related to determinants of farm sustainability ................ 169









6-2 Estimates of labor requirements for annual crop production, Cauca,
Colombia .................................................. 173

6-3 Prices of production inputs, Cauca, Colombia, December 1992 ........... 174

6-4 Fitted parameters for deterministic component of production commodity
price time series models ....................................... 175

6-5 Fitted parameters for stochastic component of production commodity price
time series models .......................................... 175

6-6 Adjustments to simulated crop yields and reported prices, Cauca, Colombia 176

6-7 Description of farm scenarios ..................................... 181

6-8 Predicted 15 year sustainability of cropping system scenarios, and McNemar
test statistic for difference from the base scenario ...................... 192

6-9 Soil loss and predicted 15 year sustainability of erosion scenarios. .......... 195

6-10 Relative sensitivity of predicted 15 year sustainability to continuous factors,
and McNemar test statistic for difference from the base scenario. .......... 197

6-11 Predicted nine-year sustainability, McNemar test statistic for difference from
the base scenario, and standard deviation of liquid assets after three years for
sources of risk scenarios .................................... 199

A-1 Object-oriented programming terms ......................... ... 211

B-1 Format of the SCENARIO section of the scenario file .................. 215

B-2 Format of the OUTPUTS section of the scenario file ................... 216

B-3 Format of the ANALYSES section of the scenario file .................. 217

B-4 Format of the RESOURCES section of the scenario file ................. 219

B-5 Format of the LINKAGES section of the scenario file .................. 221

B-6 Format of the OPERATION REQUIREMENTS section of the scenario file .. 223

B-7 A resource block within the OPERATION REQUIREMENTS section ..... 223









B-8 Format of the SCHEDULED OPERATIONS section of the scenario file .... 225

B-9 Format of the LANDSCAPE section of the scenario file ................. 226

B-10 Format of the STRATEGIES section of the scenario file ................ 228

B-11 Format of the ENTERPRISES section of the scenario file ............... 230

B-12 Format of the CONSUMPTION DECISIONS section of the scenario file ... 231

B-13 Format of the CONSTANT section of the price file .................... 232

B-12 Format of the ARMA section of the price file ......................... 233

B-12 Format of the HISTORICAL section of the price file ................... 234

C-1 Options available from FSS menu items ............................ 236















LIST OF FIGURES


Figure page


2-1 Contrasting interpretations of the relationship between chemical input levels
and sustainability ........................................... 13

3-1 Relationship between time, distribution of state and sustainability under
constant mean and high variability, and negative trend and low variability ..... 40

3-2 Estimating sustainability by sampling a small number of simulated
realizations of future system behavior ................................ 43

3-3 Sustainability and hazard of an AR(1) process with constant parameters,
declining expected value, abrupt decrease and increase in mean, decrease
and increase in variance, and decrease and increase in autocorrelation ....... 47

3-4 Simulated sustainability of a Texas rice farm under four scenarios .......... 62

4-1 Parallel ecological, economic and social hierarchies of agricultural systems .... 68

4-2 Object representation of the main components of a farming system .......... 70

4-3 Tree of resource class hierarchy in FSS ............. ................. 75

4-4 Flow of information in FSS from the generation of a field operation to its
effect on resource accounting ...................................... 89

4-5 Pseudocode representation of the USE method of the consumable resource
class ............... ............................. ........... 90

4-6 Pseudocode representation of the SELL method of the consumable resource
class ..................................................... 91

4-7 Pseudocode representation of the USE method of the timed resource class .... 92











4-8 Pseudocode representation of the USE method of the credit resource class .. .. 92

4-9 Pseudocode representation of the UPDATE method of the consumable
resource class ............... ..... .......... ............ 93

4-10 Pseudocode representation of the UPDATE method of the capital resource
class ..................................................... 93

4-11 Pseudocode representation of the UPDATE method of the credit resource
class ..................................................... 94

4-12 Pseudocode representation of the USE method of the cost list class ......... 95

4-13 Forrester representation of variable cost linkages between a consumable
resource and three linked consumable resources ........................ 96

4-14 Flow of information from the generation of a fertilizer application operation
by a crop model to its effect on the operating fund ................ ...... 97

4-15 Example of a resource box plot generated by FSS ...................... 99

4-16 Example of a final resource distribution plot generated by FSS ............ 100

4-17 Example of a sustainability time plot generated by FSS ............... 102

5-1 Location map of the study area, Cauca, Colombia .................... 107

5-2 Land use map of the Domingo farm, Cauca, Colombia ................. 108

5-3 Monthly climate statistics: mean daily solar radiation, mean daily maximum
and minimum temperature, and total rainfall, Domingo farm, Cauca,
Colombia ................................... ................ 110

5-4 Mean monthly rainfall totals at weather station used to estimate monthly
weather statistics and WGEN parameters for the Domingo farm by spatial
interpolation .......... .... .................... ......... 111

5-5 Phosphorus sorption isotherms for four soils with different mineralogy ...... 113

5-6 Decomposition of soil organic C in three allophanic and seven
non-allophanic soils ..................................... 114










5-7 Decomposition of 4C-labeled wheat straw in Lo Aguire sandy loam with
added allophane ............................................... 115

5-8 Locations and elevations of weather stations used to estimate monthly
weather statistics and WGEN parameters for the Domingo farm by spatial
interpolation ............. ............. .... ......... 123

5-9 Trend in the ratio of bean yields observed in on-farm trials and predicted
by CROPGRO, Domingo and Trujillo farms, Cauca, Colombia ........... 131

5-10 Simulated and observed bean yields, Domingo and Trujillo farms,
Cauca, Colombia, planted October 1993, March 1994, October 1994 and
M arch 1995 ............... ..................... ......... 133

5-11 Simulated and observed maize grain yields, Domingo and Trujillo farms,
Cauca, Colombia, planted October 1993, March 1994 and October 1994 .... 135

5-12 Distribution of simulated yields of October- and March-planted maize and
bean in response to historical and generated weather variability, La Florida,
Popayan, Colombia ............................................ 138

5-13 Simulated grain and biomass yield response of maize and October- and
March-planted bean applied N using the default mineralization factor ....... 140

5-14 Maize response to applied N on several volcanic soils in Narifio, Colombia .. 141

5-15 Effect of N mineralization factor, SLNF, on simulated maize response to
applied N ................... ...................... ........ 142

5-16 Simulated grain and biomass yield response of maize and October- and
March-planted bean to applied N using the adjusted mineralization factor .... 143

5-17 Grain and biomass yields of October- and March-planted bean simulated
by the original and modified versions of CROPGRO .................... 144

5-18 Grain and biomass yields of irrigated and rainfed soybean simulated by the
original and modified versions of CROPGRO ......................... 145

5-19 Effect of applied N on bean nodule growth and cumulative N2 fixed
observed and simulated with the original and modified versions of
CROPGRO, Domingo farm, Cauca, Colombia, 1994 ................... 147










5-20 Effect of applied N on bean nodule growth and cumulative N2 fixed
observed and simulated with the original and modified versions of
CROPGRO, Kuiaha, Hawaii, 1993 ................................. 148

5-21 Plant and nodule mass at 35 days, and grain yield of container-grown
beans in response to applied N .................................... 150

5-22 Simulated 60 year sequences of maize grain yields at different levels of
applied N, Domingo farm, Cauca, Colombia .......................... 151

5-23 Simulated grain yields of maize and bean in response to soil loss, Domingo
farm, Cauca, Colombia ........... ............................ 153

5-24 Simulated root distributions of October- and March-planted bean in response
to 0 cm, 20 cm, and 50 cm of soil loss, Domingo farm, Cauca, Colombia .... 155

5-25 Relative root distribution factors, F and SRGF after 0 cm, 20 cm and 50 cm
soil loss ........................................ .......... 157

5-26 Simulated grain yields of October- and March-planted bean in response to soil
loss, Domingo farm, Cauca, Colombia ............................. 159

5-27 Pseudocode representation of an algorithm for resolving resource conflicts
among crop enterprises ........................................ 160

5-28 Pseudocode representation of an algorithm for simulating erosion and crop
growth on a complex hillslope ............................... 164

6-1 Historical crop wholesale prices, Cauca, Colombia .................... 172

6-2 Cropping pattern included in farm scenario ........................... 178

6-3 Influence of type of farmer participation in on-farm trials on bean response
to ground and partially acidulated rock phosphate, chicken manure, and
10-30-10 ................ ... .. ................. .. ...... 179

6-4 Box plot of liquid assets, base scenario ........................... 187

6-5 Cumulative distribution of liquid assets after 1, 3, 6, 9, 12 and 15 years, base
scenario .................................................. 188

6-6 Sustainability and hazard time plots of the base scenario .. ............. 189









6-7 Sustainability time plot of annual cropping systems ................... 190

6-8 Sustainability time plot of base and coffee scenarios .................... 191

6-9 Sustainability time plot of base and nitrogen management scenarios ........ 193

6-10 Sustainability time plot of base and soil erosion scenarios ................ 194

6-11 Sustainability time plot of base and price scenarios ..................... 195

6-12 Sustainability time plot of base and household consumption scenarios ...... 196

6-13 Sustainability time plot of base and resource scenarios .................. 198

6-14 Sustainability time plot of base and credit scenarios .................... 199

6-15 Distribution of liquid assets after 1, 3, 6 and 9 years for the source of risk
scenarios ................................................. 200

6-16 Sustainability time plot of source of risk scenarios ................... 201

B-1 Example SCENARIO section ................. ....... ............ 215

B-2 Example OUTPUTS section ..................................... 216

B-3 Example ANALYSES section .................................... 217

B-4 Example item from the RESOURCES section ........................ 218

B-5 Example item from the LINKAGES section ...... ... ........... .... 221

B-6 Example item from the OPERATION REQUIREMENTS section ......... 224

B-7 Example item from the SCHEDULED OPERATIONS section ............ 224

B-8 Example item from the LANDSCAPE section ........................ 226

B-9 Example item from the STRATEGIES section ........................ 229

B-10 Example item from the ENTERPRISES section ....................... 229

B-11 Example item from the CONSUMPTION DECISIONS section ........... 231









D-1 Farm scenario file used to simulate the base scenario ................... 239

D-2 Crop management file used to simulate farm scenarios .................. 242

D-3 Soil profile description file used to simulate farm scenarios ............... 247


xvii














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

A SYSTEMS APPROACH TO CHARACTERIZING FARM SUSTAINABILITY

By

James William Hansen

May 1996


Chairperson: Dr. James W. Jones
Major Department: Agricultural and Biological Engineering



The potential usefulness of the concept of sustainability as a criterion for

evaluating and improving agricultural systems has been hindered by inadequacy of

approaches for its characterization. The approach presented here for characterizing farm

sustainability is based on a definition of sustainability as the ability of a system to continue

into the future, expressed by the probability that the system will continue without violating

failure thresholds during a particular future period. Characterization includes

quantification and diagnosis of constraints. Sustainability is quantified by using long-term,

stochastic simulation to sample realizations of the future behavior of a model of the

farming system. Sensitivity analysis provides a tool for testing hypotheses about

constraints to sustainability. An object-oriented farming system simulator developed for


xviii










this purpose can simulate a replicated farm scenario with stochastic inputs of weather and

prices. The farm simulator simulates the balance of farm resources and accounts for

ecological determinants of crop production by calling external crop models. Several crop

simulation models were evaluated for compatibility with a simulation study of the

sustainability of a hillside farm in the Cauca region of Colombia. The crop models were

useful for capturing response to weather variability and N management, but did not

account for important yield determinants, including P deficiency and nematode damage.

Simple modifications improved simulated response to applied N and sensitivity to soil loss.

A study of a hillside farm in Colombia showed the practical value of using simulation to

characterize sustainability. Results identified cropping system, area under cultivation,

consumption requirements, crop prices, and soil erosion as important determinants of

sustainability. The study showed that price variability contributes more than weather

variability to farm risk in this location, and that spatial diversification reduces risk and

improves sustainability. Results suggest that the farmer can enhance farm sustainability by

diversifying and intensifying crop production. Researchers can contribute to sustainability

by identifying promising high-value crops, by testing the proposed cropping systems, and

by quantifying severity and impacts of erosion. Policy makers can address sustainability

constraints caused by price volatility and lack of affordable credit.














CHAPTER 1
INTRODUCTION


The past decade has shown a shift of focus from productivity to sustainability of

agricultural systems. The growing emphasis on sustainability stems from concern about

both threats to, and negative impacts of agriculture, and from the realization that decisions

made now can have unforeseeable impacts on future generations. In spite of the growing

interest in agricultural sustainability, there is no generally accepted definition of what it is.

There is even less agreement about how to achieve it. Little progress has been made

toward developing methods for characterizing sustainability of particular agricultural

systems because of the conceptual problem of agreeing on a definition and the practical

problems that result from the fact that sustainability deals with the future and therefore

cannot be readily observed. Although some have argued that sustainability cannot and

should not be measured, there is growing realization that progress toward improving the

sustainability of agriculture is not possible without the ability to measure it and identify its

constraints.

Although concerns about agricultural sustainability focus on system levels ranging

from field to global, the farm is an appropriate level for dealing with sustainability. It is

appropriate from the standpoint of relevance; it is the farmer who makes the final

decisions about what to produce and what resources and methods will be employed. If










2

agriculture is to meet the needs of society as a whole--producing food and other products

while protecting natural resources--it must first meet the needs of the farmers who

implement and manage it. The farm level is also appropriate from a practical standpoint.

Since human goals are not intrinsic to fields or enterprises, it is difficult to identify what is

to be sustained. On the other hand, characterization of sustainability at system levels

higher than the farm is complicated by the complexity of the systems, the difficulty of

specifying system boundaries, and the often conflicting goals of multiple human actors.

The purpose of this study is to present and demonstrate a simulation-based systems

approach to characterizing sustainability of farming systems. Specific objectives of this

dissertation are as follows:

1. Review existing conceptual and methodological barriers to using the concept of
sustainability for guiding change in agriculture

2. Propose a set of elements necessary for an approach to characterizing
sustainability to provide a useful basis for guiding agriculture.

3. Present the logical and mathematical basis for a definition of sustainability that
applies generally to dynamic, hierarchical, stochastic, purposeful systems, and
relate this definition to farming systems.

4. Describe and illustrate a framework for using system simulation to quantify
sustainability and test hypotheses about its determinants.

5. Describe an object-oriented farming system simulator that was developed for the
purpose of characterizing farm sustainability, and present its data requirements.

6. Evaluate the compatibility of a set of crop models with requirements imposed by
a simulation study of sustainability in a hillside environment in Colombia.

7. Demonstrate and evaluate the approach for characterizing farm sustainability by
identifying determinants of the sustainability of a hillside farm in Colombia.












This dissertation is organized around these objectives. Chapter 2 critically reviews

existing interpretations of sustainability and approaches proposed for its characterization.

A conceptual and mathematical framework for characterizing farm sustainability is

presented in Chapter 3. Chapter 4 presents the farm simulator that was developed to

apply the proposed framework to a particular farm. The simulator uses process-level crop

simulation to characterize the contribution of ecosystem processes to farm production and

sustainability. Chapter 5 describes the physical environment in the region of Colombia

where the approach was applied, and evaluates the suitability of a set of crop models for

characterizing production and sustainability in that environment. Finally, Chapter 6

applies the approach to a particular farm in the lower Andes of southwestern Colombia.














CHAPTER 2
IS AGRICULTURAL SUSTAINABILITY A USEFUL CONCEPT?



Introduction



In literal English usage, sustainability is the ability to "keep in existence; keep up;

maintain or prolong" (Neufeldt, 1988, p. 1349). The variety of meanings acquired by

sustainability as applied to agriculture (Table 2-1) have been classified according to the

issues motivating concern (Douglass, 1984; Weil, 1990), their historical and ideological

roots (Kidd, 1992; Brklacich et al., 1991), and the hierarchical levels of systems

considered (Lowrance et al., 1986).

The distinction between sustainability as a system-describing and as a goal-

prescribing concept (Thompson, 1992) identifies two current schools-of-thought that

differ in their underlying goals. The goal-prescribing concept interprets sustainability as an

ideological or management approach to agriculture. This concept developed in response

to concerns about negative impacts of agriculture, with the underlying goal of motivating

adoption of alternative approaches. The system-describing concept interprets

sustainability either as an ability to fulfill a diverse set of goals or as an ability to continue.

This concept can be related to concerns about impacts of global change on the viability of














CHAPTER 2
IS AGRICULTURAL SUSTAINABILITY A USEFUL CONCEPT?



Introduction



In literal English usage, sustainability is the ability to "keep in existence; keep up;

maintain or prolong" (Neufeldt, 1988, p. 1349). The variety of meanings acquired by

sustainability as applied to agriculture (Table 2-1) have been classified according to the

issues motivating concern (Douglass, 1984; Weil, 1990), their historical and ideological

roots (Kidd, 1992; Brklacich et al., 1991), and the hierarchical levels of systems

considered (Lowrance et al., 1986).

The distinction between sustainability as a system-describing and as a goal-

prescribing concept (Thompson, 1992) identifies two current schools-of-thought that

differ in their underlying goals. The goal-prescribing concept interprets sustainability as an

ideological or management approach to agriculture. This concept developed in response

to concerns about negative impacts of agriculture, with the underlying goal of motivating

adoption of alternative approaches. The system-describing concept interprets

sustainability either as an ability to fulfill a diverse set of goals or as an ability to continue.

This concept can be related to concerns about impacts of global change on the viability of











5

Table 2-1. Interpretations of agricultural sustainability.
Sustainability as an ideology:

". .. a philosophy and system of farming. It has its roots in a set of values that reflect a state of empowerment, of awareness
of ecological and social realities, and of one's ability to take effective action." (MacRae et al., 1990, p. 156)

". .. an approach or a philosophy that integrates land stewardship with agriculture. Land stewardship is the philosophy
that land is managed with respect for use by future generations." (Neher, 1992, p. 54)

"... a philosophy based on human goals and on understanding the long-term impact of our activities on the environment
and on other species. Use of this philosophy guides our application of prior experience and the latest scientific advances to
create integrated, resource-conserving, equitable farming systems." (Francis & Youngberg, 1990, p. 8)

"... farming in the image of Nature and predicated on the spiritual and practical notions and ethical dimensions of
responsible stewardship and sustainable production of wholesome food." (Bidwell, 1986, p. 317)

Sustainability as a set ofstrategies:

"... a management strategy which helps the producers to choose hybrids and varieties, a soil fertility package, a pest
management approach, a tillage system, and a crop rotation to reduce costs of purchased inputs, minimize the impact of the
system on the immediate and the off-farm environment, and provide a sustained level of production and profit from
farming." (Francis, Sander & Martin, 1987, p. 12)

"... a loosely defined term for a range of strategies to cope with several agriculturally related problems causing increased
concern in the U.S. and around the world." (Lockeretz, 1988, p. 174)

Farming systems are sustainable if "they minimize the use of external inputs and maximize the use of internal inputs already
existing on the farm." (Carter, 1989, p. 16)

"... (a) the development of technology and practices that maintain and/or enhance the quality of land and water resources,
and (b) the improvements in plants and animals and the advances in production practices that will facilitate the substitution
of biological technology for chemical technology." (Ruttan, 1988, p. 129)

Sustainability as the ability tofulfill a set of goals:
"A sustainable agriculture is one that, over the long term, enhances environmental quality and the resource base on which
agriculture depends, provides for basic human food and fiber needs, is economically viable, and enhances the quality of life
for farmers and society as a whole." (American Society of Agronomy, 1989, p. 15):

". .. agricultural systems that are environmentally sound, profitable, and productive and that maintain the social fabric of the
rural community." (Keeney, 1989, p. 102)

"... an agrifood sector that over the long term can simultaneously (1) maintain or enhance environmental quality, (2)
provide adequate economic and social rewards to all individuals and firms in the production system, and (3) produce a
sufficient and accessible food supply." (Brklacich, Bryant & Smit, 1991, p. 10)

". an agriculture that can evolve indefinitely toward greater human utility, greater efficiency of resource use, and a
balance with the environment that is favorable both to humans and to most other species." (Harwood, 1990, p. 4)

Sustainability as the ability to continue:

"A system is sustainable over a defined period if outputs do not decrease when inputs are not increased." (Monteith, 1990,
p. 91)
"... .the ability of a system to maintain productivity in spite of a major disturbance, such as is caused by intensive stress or a
large perturbation." (Conway, 1985, p. 35)
"... the maintenance of the net benefits agriculture provides to society for present and future generations." (Gray, 1991, p.
628)

"Agriculture is sustainable when it remains the dominant land use over time and the resource base can continually support
production at levels needed for profitability (cash economy) or survival (subsistence economy)." (Hamblin, 1992, p. 90)










6

agriculture, and to the goal of using sustainability as a criterion for guiding agriculture as it

responds to rapid changes in its physical, social and economic environment.

Although the concept of sustainability has been useful for consolidating concerns

and motivating change, concrete examples of its use as an operational criterion for guiding

efforts to improve agricultural systems are difficult to identify. The objectives of this

chapter are (1) to examine conceptual and methodological barriers to using the concept of

sustainability for guiding change in agriculture, and (2) to propose a set of elements

necessary for an approach to characterizing sustainability to provide a useful criterion for

guiding agriculture.



Sustainability as an Approach to Agriculture



The sustainable agriculture movement evolved from several reform movements in

the U.S.A., Canada and Western Europe that developed in response to concerns about

impacts of agriculture such as depletion of nonrenewable resources, soil degradation,

health and environmental effects of agricultural chemicals, inequity, declining rural

communities, loss of traditional agrarian values, food quality, farm worker safety, decline

in self-sufficiency, and decreasing number and increasing size of farms. These problems

became associated with "conventional agriculture" that was perceived as unsustainable

(Dahlberg, 1991). "Alternative agriculture" is often equated with sustainable agriculture

(O'Connell, 1992; Madden, 1987; Harwood, 1990; Dahlberg, 1991; Bidwell, 1986) and

reflects the goal of promoting alternatives to conventional agriculture. Reviews by












Harwood (1990) and Kidd (1992) trace the historical development of the sustainable, or

alternative agriculture movement.

Differences in values and practices promoted as sustainable have been attributed to

differences in the problems emphasized (Carter, 1989) and to different visions of what

agriculture should be like (Thompson, 1992). "Originally, the advocates of alternative

approaches to agriculture--all united in their critique of industrial agriculture as being

unsustainable--debated among themselves the future direction and shape of agriculture"

(Dahlberg, 1991, p. 337). Some have focused on identifying sustainable alternatives to

existing management practices while others have advocated new philosophical orientations

toward agriculture.


Sustainability as an alternative ideology


MacRae et al. (1990), Neher (1992) and Francis and Youngberg (1990) defined

sustainable agriculture as a philosophy (Table 2-1). Ikerd (1991) described low-input,

sustainable agriculture (LISA) as more a philosophy than a practice. Examining the

concept of conventional agriculture is important since sustainable agriculture is often

described by its contrast with conventional agriculture (Lockeretz, 1988; MacRae et al.,

1989; Hauptli et a., 1990; Dobbs et al., 1991; O'Connell, 1992; Hill and MacRae, 1988).

Conventional agriculture. The concept of conventional agriculture was apparently

developed in order to clarify, and justify alternative approaches to agriculture.

Conventional agriculture is characterized as "capital-intensive, large-scale, highly

mechanized agriculture with monocultures of crops and extensive use of artificial












fertilizers, herbicides and pesticides, with intensive animal husbandry" (Knorr and

Watkins, 1984, p. x) with a paradigm of "strength through exhaustion" (Bidwell, 1986, p.

317). Hill and MacRae (1988) contrasted approaches of conventional and sustainable

agriculture (Table 2-2). Based on a review of actual cropping practices in the U.S.,

Madden (1990) appropriately identified conventional agriculture as a caricature.

Beus and Dunlap (1990) identified centralization, dependence, competition,

domination of nature, specialization, and exploitation as key elements of conventional

agriculture from the writings of six conventional agriculture advocates. Although

"conventional agriculture" was applied to mainstream U.S. agriculture as one side of a

debate between competing paradigms, its description was admittedly a construct for the

purpose of"clarifying opposing positions, facilitating comparisons, and sharpening the

focus of the debate" (p. 597). A survey of agriculturalists in Washington state by the same

authors (Beus and Dunlap, 1991) suggested that the conventional agriculture that they

described does not represent mainstream U.S. agriculture. A random sample of farmers

and three of the four groups of conventional agriculturalists surveyed showed greater

agreement with the alternative than with the conventional agriculture paradigm.

Characterization of conventional agriculture extends to mainstream research and

education institutions where research has been described as too narrow, short-sighted,

biased by interests of agribusiness funding sources, and distorted by the values of scientists

to be able to deal with the issues necessary to achieve sustainability (Bidwell, 1986; Allen

and van Dusen, 1988; MacRae et al., 1989; Dahlberg, 1991; Kirschenmann, 1991; Hill and

MacRae, 1988). Beus and Dunlap (1990) and Dahlberg (1991) expressed concern that














Table 2-2. Contrasting approaches of conventional and sustainable agriculture as
characterized by Hill and MacRae (1988, p. 95).


Conventional agriculture


Symptoms

Reductionist

Eliminate "Enemies"

Narrow focus (neglects side effects, health
& environmental costs ignored)

Instant

Single, simple (magic bullet, single discipline)


Temporary solutions

Unexpected disbenefits (to person & planet)

High power (risk of overkill & errors/ accidents)

Direct "attack"


Imported

Products

Physico-chemical (often unnatural, synthetic)

Technology-intensive

Centralized

Values secondary

Expert, paternalistic (arrogant)

Dependent

Inflexible

Ignores freedom of choice (unjust)

Disempowering

Competitive

Authored


Sustainable agriculture


Causes, prevention

Holistic

Respond to indicators

Broad focus (subcellular to all life to globe, all
costs internalized)

Long time frame (future generations)

Multifaceted, complex (multi- & trans-
disciplinary)

Permanent solutions

Unexpected benefits

Low power (minimal risk)

Indirect, benign approaches (catalytic, multiplier,
synergistic effects)

Local solutions and materials

Processes, services

Bio-ecological (natural)

Knowledge/skill intensive

Decentralized (human scale)

Compatible with higher values

Individual/community responsibility (humble)

Self-maintaining/regulating

Flexible

Respects freedom of choice (just)

Empowering

Co-operative

Anonymous (seeking neither reward or fame)












such institutions threaten to dilute the concept of sustainable agriculture by co-opting it

while ignoring its more important and radical aspects.

Alternative values. Sustainable agriculture has been described as an umbrella term

encompassing several ideological approaches to agriculture (Gips, 1988) including organic

farming, biological agriculture, alternative agriculture, ecological agriculture, low-input

agriculture, biodynamic agriculture, regenerative agriculture, permaculture, and

agroecology (Carter, 1989; MacRae et al., 1989; Bidwell, 1986; O'Connell, 1992;

Kirschenmann, 1991; Dahlberg, 1991).

Beus and Dunlap (1990) listed decentralization, independence, community,

harmony with nature, diversity, and restraint as key values of alternative agriculture.

Social values such as equity, the value of traditional agricultural systems, self-sufficiency,

preservation of agrarian culture, and preference for small, owner-operated farms have

been incorporated into definitions of sustainability (Weil, 1990; Keeney, 1989; Bidwell,

1986; Francis and Youngberg, 1990). The concept of equity is extended to include future

generations (Batie, 1989; Norgaard, 1991). Environmental values associated with

sustainability include mimicry of nature and an "ecocentric" ethic. Hauptli et al. (1990)

described mimicry of nature: ". sustainable agriculture attempts to mimic the key

characteristics of a natural ecosystem .. (p. 143). The ecocentric position--valuing

ecosystems or species without regard to their impact on human welfare--is illustrated by

Douglass (1984) who stated ecology-minded people ". define agricultural sustainability

in biophysical terms, and to allow its measurement to determine desirable population

levels" (p. 5).












Sustainability as a set of strategies


Francis and Youngberg (1990) described sustainable agriculture as a philosophy

that guides the creation of farming systems. Specific management strategies are often

suggested by ideological interpretations of sustainability. The strategies promoted as

sustainable (Table 2-3) are based on the types of problems emphasized and on views of

what would constitute an improvement.


Table 2-3. Strategies frequently associated with sustainability.
Strategy References

Self-sufficiency through preferred use of on-farm or locally a, b, g, d
available "internal" resources to purchased "external" resources.

Reduced use or elimination of soluble or synthetic fertilizers, a, e, f, h, i, d, k

Reduced use or elimination of chemical pesticides, substituting a, c, d, e, f, h, i, j, k
integrated pest management practices.

Increased or improved use of crop rotations for diversification, a, c, d, e, f, h, j
soil fertility, and pest control.

Increased or improved use of manures and other organic a, c, f, h, j, k
materials as soil amendments.

Increased diversity of crop (and animal) species. a, d, g, i

Maintenance of crop or residue cover on the soil. a, d, e

Reduced stocking rates for animals, a, c, d
SLockeretz, 1988
bHarwood, 1990
" MacRae et al. 1990
dNeher, 1992
" Dobbs et al. 1991
'MacRae et al. 1989
SGliessman, 1990
h Edwards, 1990
'Hauptli et al. 1990
SO'Connell, 1992
k Hill & MacRae, 1988










12

The strategy most frequently linked to sustainability is reduction or elimination of

the use of processed chemicals, particularly fertilizers and pesticides (Stinner and House,

1987; Lockeretz, 1988; Carter, 1989; Hauptli et al., 1990; Madden, 1990; Dobbs et al.,

1991). In 1988, the U.S. Department of Agriculture linked sustainability to levels of

inputs by establishing the LISA (low-input, sustainable agriculture) research program

(O'Connell, 1990; Dicks, 1992). Arguments for reducing chemical inputs include limited

supplies of fossil fuels, decreasing commodity prices necessitating reducing input costs, a

need for self-sufficiency, concerns about pollution, and health and safety concerns (Francis

and King, 1988; Carter, 1989; Stinner and House, 1989; Conway and Barbier, 1990;

MacRae et al., 1990; Rodale, 1990).

York (1991) argued that fewer options exist for reducing fertilizer inputs than

pesticide inputs in agricultural systems while maintaining sustainable production. Unlike

pesticides, soil nutrient elements generally have no substitutes and are subjected to harvest

and other losses that must be replaced by weathering or imported from outside the system

if production is to be sustained. The high energy cost unique to N fertilizer production

and the potential for biological fixation suggest a need and potential for seeking

alternatives to synthetic N fertilizers that does not exist for mineral-based nutrients such as

P and K.

The important distinction between production systems that currently employ high

levels of chemical inputs and those that employ low levels (Weil, 1990) is often

overlooked. Zandstra (1994) described sustainability as a function of chemical input levels

(Fig. 2-la). Excessive input levels were said to degrade natural resources through












A

exhaustion


accumulaion


input level


B
/\
/
//




S / \
OT /
inputs control













sustainability


Figure 2.1. Contrasting interpretations of the relationship between chemical input
levels and sustainability by (a) Zandstra (1994) and by (b) Stinner and House (1987).










14

accumulation while inadequate levels degrade resources through exhaustion. This concept

is in sharp contrast to the decreasing relationship between chemical input levels and

sustainability proposed by Stinner and House (1987) (Fig. 2-1b).

Studies in Mali, Benin, Zambia, and Tanzania provide examples of resource

degradation due to inadequate chemical inputs (Budelman and van der Pol, 1992). In each

case, supplies of soil nutrients were exhausted rapidly due to a combination of harvest,

erosion, leaching, denitrification and volatilization, with harvest being the greatest loss.

Nutrient budgets estimated for several crops in southern Mali were always negative for N

and K, and variable but generally better for P. The authors concluded that the only way to

make these cropping systems sustainable is with increased use of fertilizers.


Discussion


Interpreting sustainability as an approach to agriculture has been useful for

motivating change. Sustainability as an ideology has provided as a common banner for

various agricultural reform movements (Gips, 1988; Dahlberg, 1991). Research and

promotion of sustainability interpreted as a set of strategies has become part of policy in

the U.S. in the form of provisions in the 1990 Farm Bill (O'Connell, 1992; Yetley, 1992).

Interpreting sustainability as an approach is not useful for guiding change in

agriculture for several reasons. First, approaches developed in response to problems in

North America and Europe may be inappropriate in regions where circumstances and

problems are different. The alternative agriculture movement has its roots primarily in

regions characterized by high levels of resource consumption, food surpluses, high levels












of chemical inputs, relatively deep, fertile soils, and relatively stable populations. In

contrast, many less developed tropical regions are characterized by lower levels of

resource consumption, frequent or chronic food shortages, lower levels of chemical

inputs, relatively fragile soils, and rapidly growing populations. Attempts to link strategies

to sustainability by definition fail to consider the need to match technologies to specific

environments.

Dicks (1992) argued that interpretations of sustainability in the U.S. have been

shaped by food surpluses. A shift in concern from global food security to environmental

quality in the 1980s (York, 1991) led to the perspective that ". the question is not can

we produce more food, but what are the ecological consequences of doing so?"

(Douglass, 1984, p. 5). However, much of the concern about sustainability in less

developed countries is related to the need to increase productivity to meet future needs of

growing populations (Ruttan, 1988; York, 1988, 1991; Lynam and Herdt, 1989;

Plucknett, 1990). The potential for the desperation imposed by poverty to shorten

people's planning horizons (Ashby, 1985) raises questions about the ecological

consequences of failing to produce more food (Mellor, 1988; Oram, 1988). The

alternative agriculture movement has not adequately addressed the need to feed rapidly

growing populations in order to prevent both human and ecological disaster.

The second problem is that a distorted caricature of conventional agriculture may

cause approaches that may enhance sustainability to be ignored or rejected because of

their association with conventional agricultural institutions. Although the philosophical

roots of the alternative agriculture movement formed outside of the academic community










16

(Rodale, 1990; Bidwell, 1986), most of the practices it promotes as sustainable are largely

products of mainstream research and educational institutions (Francis and Sahs, 1988;

York, 1988).

Third, establishing the contribution of an approach to sustainability through

definition eliminates the perceived need to evaluate approaches that may be poor or

harmful in a particular context. If strategies are identified as sustainable based on their

effect on agricultural systems, and agricultural systems are then judged to be sustainable

based on their implementation of sustainable strategies, then a form of circular logic

results. It is logically impossible to evaluate the contribution of an approach to

sustainability when adherence to that approach has already been used as a criterion for

evaluating sustainability. This circular logic is a fourth reason why interpreting

sustainability as an approach is not useful for guiding change.

Because of the temporal nature of sustainability, errors of either ignoring

approaches that enhance sustainability or promoting approaches that threaten it may not

be obvious when the approaches are implemented. The evaluation needed to recognize

errors and improve approaches is not possible if sustainability is interpreted as a

philosophy or a set of strategies. Thompson (1992) warned that

Our society may collapse because of shortsighted stupidity on the part of
pro-growth, resource exploiting power elites, but the collapse will only be
tragic if it is shortsightedness or ignorance on the part of environmentally
and ethically concerned people that helps bring it about. (p. 19)













Sustainability as a Property of Agriculture



The concept of sustainability as an approach to agriculture evolved in parallel with

the concept of sustainability as a system property. While Dahlberg (1991) argued that

"sustainability" was first used by an emerging alternative agriculture movement to

prescribe a particular set of values, Kidd (1992) countered that the system describing

concept developed earlier but did not use the word "sustainability" until later. As a

property of agriculture, sustainability is interpreted as either the ability to satisfy a diverse

set of goals or an ability to continue through time.


Sustainability as an ability to satisfy goals


A sustainable agricultural system is often defined as one that fulfills a balance of

several goals through time. These goals generally include some expression of maintenance

or enhancement of the natural environment, provision of human food needs, economic

viability, and social welfare (Table 2-1).

Lynam and Herdt (1989) argued that an interpretation of sustainability based on

several qualitative goals fails to provide a criterion useful for guiding agricultural research.

Ifa system is defined as sustainable when it protects the natural environment, provides

adequate food, and maintains producer profitability, then there is no logical way to rank,

for example, the relative importance of commodity price variability and nitrate leaching

into aquifers as determinants of sustainability. Furthermore, the subjectivity of goal













Sustainability as a Property of Agriculture



The concept of sustainability as an approach to agriculture evolved in parallel with

the concept of sustainability as a system property. While Dahlberg (1991) argued that

"sustainability" was first used by an emerging alternative agriculture movement to

prescribe a particular set of values, Kidd (1992) countered that the system describing

concept developed earlier but did not use the word "sustainability" until later. As a

property of agriculture, sustainability is interpreted as either the ability to satisfy a diverse

set of goals or an ability to continue through time.


Sustainability as an ability to satisfy goals


A sustainable agricultural system is often defined as one that fulfills a balance of

several goals through time. These goals generally include some expression of maintenance

or enhancement of the natural environment, provision of human food needs, economic

viability, and social welfare (Table 2-1).

Lynam and Herdt (1989) argued that an interpretation of sustainability based on

several qualitative goals fails to provide a criterion useful for guiding agricultural research.

Ifa system is defined as sustainable when it protects the natural environment, provides

adequate food, and maintains producer profitability, then there is no logical way to rank,

for example, the relative importance of commodity price variability and nitrate leaching

into aquifers as determinants of sustainability. Furthermore, the subjectivity of goal










specification links criteria for determining sustainability to the goals and values of the

analyst or the author of a definition rather than to the agricultural system. At the farm

level and higher, goals belong to the actors within the system and are, therefore,

endogenous. Kidd (1992) argued that it is not helpful to "use sustainability loosely as a

general purpose code word encompassing all of the aspects of agricultural policy that the

authors consider desirable" (p. 24).


Sustainability as an ability to continue


The final concept interprets sustainability as a system's ability to continue through

time. Hildebrand (1990) suggested that sustainability may be interpreted as the length of

time that a system can be maintained. According to Hamblin (1992), sustainability implies

that agriculture remains the dominant land use. Lynam and Herdt (1989) and Jodha

(1990) expressed sustainability in terms of maintaining some level of output. Monteith

(1990) added consideration of the possible confounding interaction of changes in input

and output levels. The definitions of Fox (1991) and Hamblin (1992) emphasized the

continuing ability to meet human needs. Conway (1985), Conway and Barbier (1990) and

Altieri (1987) emphasized the ability to withstand disturbances.

Interpreting sustainability as an ability to continue is consistent with literal English

usage of "sustain" and its derivatives. Its potential usefulness comes from suggesting

criteria for characterizing sustainability, providing a basis for identifying constraints and

evaluating proposed approaches to its improvement. This potential usefulness has been

limited by inadequacy of current approaches for characterizing sustainability.













Approaches to Characterizing Sustainability



Characterization is a prerequisite to applying the concept of sustainability as a

criterion for identifying constraints, focusing research, and evaluating and improving

agricultural policy and practices. The conceptual problem of defining sustainability and

methodological problems imposed by its temporal nature have hindered development of

approaches to characterizing sustainability. Sustainability involves future outcomes that

cannot be observed in the time-frame required for intervention (Lynam and Herdt, 1989;

Harrington, 1992). For this reason, Conway (1994) argued that defining sustainability in

terms of preservation or duration has little practical value.

The variety of approaches reviewed here reflects the different interpretations of

sustainability and methodological difficulties that result from its temporal nature.

Characterization by adherence to prescribed approaches is based on an interpretation of

sustainability as an approach to agriculture. Characterization by multiple qualitative

indicators and attempts to integrate such indicators are consistent with interpreting

sustainability as an ability to satisfy diverse goals. Sustainability as an ability to continue is

usually characterized by time trends or resilience.


Adherence to prescribed approaches


In a study comparing conventional and sustainable farms in South Dakota, Dobbs

et al. (1991) identified farms as sustainable if they reduced chemical inputs relative to













Approaches to Characterizing Sustainability



Characterization is a prerequisite to applying the concept of sustainability as a

criterion for identifying constraints, focusing research, and evaluating and improving

agricultural policy and practices. The conceptual problem of defining sustainability and

methodological problems imposed by its temporal nature have hindered development of

approaches to characterizing sustainability. Sustainability involves future outcomes that

cannot be observed in the time-frame required for intervention (Lynam and Herdt, 1989;

Harrington, 1992). For this reason, Conway (1994) argued that defining sustainability in

terms of preservation or duration has little practical value.

The variety of approaches reviewed here reflects the different interpretations of

sustainability and methodological difficulties that result from its temporal nature.

Characterization by adherence to prescribed approaches is based on an interpretation of

sustainability as an approach to agriculture. Characterization by multiple qualitative

indicators and attempts to integrate such indicators are consistent with interpreting

sustainability as an ability to satisfy diverse goals. Sustainability as an ability to continue is

usually characterized by time trends or resilience.


Adherence to prescribed approaches


In a study comparing conventional and sustainable farms in South Dakota, Dobbs

et al. (1991) identified farms as sustainable if they reduced chemical inputs relative to










20

typical farms, and included rotations, legumes, tillage and cover crops for management of

fertility, erosion and weeds. Sustainable farms were sampled by sending questionnaires to

farmers they ". believed might be using greatly reduced or even zero levels of synthetic

chemicals in their farming operations" (p. 111). Cordray et al. (1993) characterized the

sustainability of farmers in Washington and Oregon based on changes in agricultural

chemical use and adoption of alternative production practices. Taylor et al. (1993)

developed a quantitative index of sustainability based on production practices of

Malaysian cabbage farmers. Practices were assigned values according to their "inherent

sustainability" determined by consensus of the research team, weighted by their expected

contribution to sustainability, then combined into a composite index evaluated for each

farm.

Taylor et al. (1993) were the only authors in the above studies to acknowledge the

necessity of assuming a relationship between farmer practices and future viability if

practices are to serve as a basis for characterizing sustainability. Even if evidence

supporting such a relationship were presented, circular logic prevents using sustainability

measured by adoption of particular practices as a criterion for evaluating and improving

agricultural practices.


Multiple qualitative indicators


Torquebiau (1992) used a set of indicators to characterize sustainability of tropical

agroforestry home gardens. Several system attributes that were believed to influence

sustainability were identified related to the resource base, system performance, and effects












on other systems. Measurable indicators were identified for each system attribute. A

negative change in an individual indicator indicated unsustainability. Jodha (1990)

developed a set of indicators of unsustainability of mountain agriculture in the Himalayan

region consisting of visible changes in natural resources and farming practices that indicate

system degradation, changes in farming practices that compensate for less visible changes

in the environment, and inappropriate development initiatives that may lead to negative

impacts. Neher (1992) described an approach that is being developed to monitor

agroecosystem health at regional and national scales in the U.S. A number of indicators of

environmental quality and agricultural performance are to be measured to provide a

baseline, then monitored to identify changes.

Monitoring sets of qualitative indicators is consistent with interpreting

sustainability as the ability to meet a diverse set of goals and the belief that no single

indicator can exist (Geng et al., 1990; Norgaard, 1991). However, diverse sets of

indicators are difficult to interpret and do not provide mechanisms for diagnosing causes

of unsustainability, or for evaluating effects of proposed interventions. Such indicators do

not facilitate establishing cause-effect relationships between diverse system properties.


Integrated, quantitative indicators


Increasing recognition of the need for quantification has motivated efforts to

combine diverse indicators.of sustainability into integrated, quantitative measures. One

goal of the project described by Neher (1992) for monitoring agroecosystem health is to

combine indicators into an aggregate measure of agricultural sustainability in a manner












that balances productivity, environmental soundness and socioeconomic viability goals.

Lal (1991) proposed a sustainability coefficient as a function of output per unit of input at

optimal per capital productivity or profit, output per unit of decline in the most limiting or

least renewable resource, and the minimum assured output level. Sands and Podmore

(1993) proposed an environmental sustainability index as an aggregation of sub-indices of

soil productivity, ecosystem stability, and potential to degrade the environment. Selection

of components of the sub-indices and the form of aggregation functions were indicated as

important research topics. Stockle et al. (1994) proposed a framework for evaluating

sustainability based on nine system attributes: profitability, productivity, quality of soil,

water, and air, energy efficiency, fish and wildlife habitat, quality of life, and social

acceptance. Production system sustainability is determined by scoring attributes as

weighted functions of quantifiable, long-term constraints, then combining weighted

attribute scores into an integrated measure.

The consistent inability to specify aggregation functions in these studies points to

the weakness of interpreting sustainability as the ability to fulfill diverse sets of goals as a

conceptual foundation for characterization. Diagnosis is limited by need to decide a-priori

the relative importance of different types of constraints to sustainability. Stockle et al.

(1994) acknowledged and defended the subjectivity needed to aggregate diverse system

attributes into an integrated measure of sustainability.












Time trends


Time trend approaches express sustainability in terms of the direction and degree

of measurable changes in system properties through time. Lynam and Herdt (1989)

regarded a system as sustainable if there is a non-negative trend in its output. They

proposed total factor productivity, the total value of system outputs divided by the value

of system inputs, as the output criterion because it accounts for changes in the value of

inputs. Ehui and Spencer (1992) extended total factor productivity to account for changes

in the value of natural resource stocks, particularly soil nutrients. Hedgerow intercropping

in Nigeria was determined to be unsustainable without correction for soil nutrient flows,

but sustainable after accounting for nutrients. Monteith (1990) proposed determining

sustainability from a contingency table of trends of inputs and outputs (Table 2-4). Cereal

production was determined to be sustainable in the Karimnagar district in Andhra Pradesh,

India, based on increasing yields and decreasing land use during 27 years. Decreases in

both land use and yields in the Adilabad district prevented inference about sustainability.



Table 2-4. Contingency table for inferring sustainability based on trends of
system inputs and outputs (Monteith, 1990).
--------------------Inputs-------------------------

Outputs decreasing constant increasing

decreasing indeterminant unsustainable unsustainable
constant sustainable sustainable unsustainable

increasing sustainable sustainable indeterminant












Characterizing sustainability by time trends is appealing because of its simplicity.

The slope of the estimated trend line provides a quantitative index with an intuitive

interpretation as a rate of system deterioration or enhancement. Trends represent an

aggregate response to several determinants of sustainability, eliminating the need to devise

and defend aggregation schemes.

The assumption needed to infer sustainability from trends--that future rates of

system degradation can be approximated by past rates--is often difficult to defend.

Unsustainability can express itself either as a gradual change or as an abrupt collapse

(Conway, 1985; Trenbath et al., 1990). Furthermore, much of the concern about

sustainability comes from recognition that agriculture is being impacted by unprecedented

changes in population pressure, resource demands, market structures, and technology.

Another weakness is the manner in which time-trend approaches interpret temporal

variability. Variability tends to hinder sustainability by driving subsistence farmers to

desperation, leading to environmental degradation that may not recover during normal or

good periods (Mellor, 1988). Price and yield variability have also been shown to increase

the probability of farm failure in the U.S. (Grant et al., 1984; Perry et al., 1986).

However, when characterization is based on time trends, either variability is ignored or it

implicitly enhances sustainability by reducing the probability of identifying a significant

negative trend.

A final criticism is that applications of time trends to sustainability have examined

levels of system performance without considering the levels of needs and goals of the

individuals or segments of society who decide on the fate of those systems.












Resilience


Conway (1985) defined sustainability as resilience: ". the ability of a system to

maintain productivity in spite of a major disturbance" (p. 25). He suggested that

measurement of five system properties are necessary to characterize resilience: inertia,

elasticity, amplitude, hysteresis, and malleability (Conway, 1994). Cramb (1993) based

inferences about the sustainability of two shifting cultivation systems in eastern Malaysia

on both trends and resilience. Pepper production was determined to be sustainable

because production in 1989 recovered to its 1980 level in response to price recovery after

production diminished (in 1985) due to a period of low prices. Rubber production was

considered more sustainable at Batu Linang than at Nanga Tapih, as indicated by recovery

of depressed production in response to price recovery.

Like time trends, resilience can be viewed as an aggregate system response to

determinants of sustainability. However, inability to identify a single measure of resilience

(Conway, 1994) leads to the same problems of interpretation faced when using a diverse

set of indicators to characterize sustainability. Assumptions about the likelihood and

timing of disturbances have been avoided by interpreting sustainability as an intrinsic

property of an agricultural system in isolation from its environment (Conway and Barbier,

1990). However, York (1988) argued that (un)sustainability is not an intrinsic property

but rather a response to changing environmental and socioeconomic circumstances.

Predictions about future sustainability cannot be made in the absence of assumptions about

changes and variability in those higher-level systems that comprise a system's










26

environment. Resilience shares with time trend approaches the criticism that it ignores the

goals of the human actors within agricultural systems.


System simulation


Simulation has been used to characterize the sustainability of crop production in

response to soil dynamics. Singh and Thornton (1992) illustrated the use of long-term

simulation of crop sequences replicated with stochastic inputs of weather data to examine

trends and variability in yields. Lerohl (1991) used the Erosion-Productivity Impact

Calculator (EPIC) (Williams et al., 1984) to study the long-term impact of predicted soil

erosion on productivity of crop rotations on four soil types in Alberta, Canada.

Sustainability was inferred in all soil-rotation combinations because no negative trend in

crop yields could be detected during a simulated 100 year period.

Other studies have used crop simulation models to examine relationships between

production and environmental degradation. Singh and Thornton (1992) illustrated the use

of CERES-Maize (Jones and Kiniry, 1986) to simulate the effects of soil type and rate of

application of N fertilizer on distributions of maize grain yield and NO, leaching into

groundwater from upland fields in Chiang Mai, Thailand. Alocilja and Ritchie (1993) used

CERES-Maize and a multiple goal optimization technique to identify sets of N fertilization

schedules that were optimal in the sense that neither production nor water quality could be

improved without decreasing satisfaction of the other goal.

Several whole-farm simulation studies have looked at the effect of various factors

on farm survivability. For example, Perry et al. (1986) examined effects of production












costs, labor availability, rice grain quality, land tenure, trends and variability of rice and

soybean prices and yields, beginning equity, type of rotations, and participation and terms

of government farm programs on probability of farm survival during a five year period.

Production and environmental processes were not simulated. Although this and similar

studies have not generally used the word "sustainability," the use of farm survival as a

criterion is consistent with an interpretation of sustainability as economic viability

(Madden, 1987; Lockeretz, 1988; Dicks, 1992; Neher, 1992). Survivability addresses

shortcomings of other approaches by integrating levels, trends and variability in system

performance with the needs and goals of farmers.

System simulation is a tool, and does not suggest a particular criterion for

evaluating sustainability. Simulation can be used to examine future impacts of alternative

interventions across the range of expected variability in a manner that is not possible with

empirical observation and experimentation. The value of simulation is limited by

capabilities of, and confidence in simulation models, by availability and reliability of input

data, and by lack of methods for designing and interpreting simulation studies for

characterizing sustainability. So far, there has been little integration of models of crop and

animal production, environmental degradation, economic processes, and farmer decisions.



Elements of a Useful Approach for Characterizing Sustainability



In order for sustainability to be a useful criterion for guiding change in agriculture,

several elements should be incorporated into approaches to its characterization (Table 2-










28

5). First, characterization should be based on a literal interpretation of sustainability (Fox,

1991). Regardless of the merits of goals and ideals frequently incorporated into

definitions of sustainability, if the idea of continuation through time is omitted then those

ideals and goals are something other than sustainability.


Table 2-5. Elements of a useful approach to characterizing sustainability of agricultural
systems.
Element Explanation

Literal Defines sustainability as an ability to continue through time, consistent
with literal English usage.

System- Identifies sustainability as an objective property of a particular
oriented agricultural system whose components, boundaries, and context in
hierarchy are clearly specified.
Quantitative Treats sustainability as a continuous quantity, permitting comparisons of
alternative systems or approaches.

Predictive Deals with the future rather than the past or present.

Stochastic Treats variability as a determinant of sustainability and a component of
predictions.
Diagnostic Uses an integrated measure of sustainability to identify and prioritize
constraints.


Second, characterization should be system-oriented. A literal interpretation

suggests that sustainability is an objective property of an agricultural system. It cannot be

a property of approaches to agriculture if it is to serve as a basis for evaluating and

improving approaches. Lynam and Herdt (1989) argued that sustainability is a relevant

criterion for evaluating technology only when the system is clearly specified, including its

boundaries, components, and context in hierarchy. Sustainability has meaning only in the

context of specific temporal and spatial scales. Fresco and Kroonenberg (1992) cited a












number of examples in which disturbances that threaten sustainability at one spatial and

temporal scale could be seen as natural cycles at broader scales. Both constraints to

sustainability and factors that can be managed for its enhancement depend on the level of

the system (Spencer and Swift, 1992). The objectivity that results from a system-oriented

approach is essential for guiding change, but may work against motivating change because

it may call prescribed approaches into question.

Third, an approach to characterizing sustainability should be quantitative.

Although MacRae et al. (1989) cited quantification as a barrier to sustainability, others see

it as a prerequisite to using sustainability as a criterion for evaluating and improving

agricultural systems (Monteith, 1990; Harrington, 1992). Sustainability is often treated as

a discrete property: "A farm is either sustainable or it's not sustainable. Simply by

definition, you cannot create a system that is half sustainable" (Rodale, 1990, p. 273).

However, comparisons among agricultural systems or alternative approaches are possible

only when sustainability is treated as a continuous quantity.

Fourth, since sustainability deals with future changes, its characterization must be

predictive of the future rather than merely descriptive of the past or present (Harrington,

1992). Sustainability has little meaning after the fact. The deterministic view that ". .. a

farm will either last for a very long period, or it won't" (Rodale, 1990, p. 273) does not

take into account the uncertainty of predictions resulting from the inherent variability of

the farming system's environment. A stochastic approach, the fifth element, recognizes

variability as a determinant of sustainability and appropriately expresses predictions in

terms of probabilities.









30

Finally, characterization of sustainability should be diagnostic. Sustainability is a

useful concept when its characterization focuses research and intervention by identifying

and prioritizing constraints. Diagnosis can be accomplished by testing hypotheses about

constraints based on a measure of sustainability that is both comprehensive and integrated.

Diagnosis is facilitated by use of a single measure of sustainability that combines the range

of possible determinants into a single, integrated measure of system response. An

integrated measure is necessary for comparing, for example, the relative impact of nitrate

leaching into aquifers and product price volatility on sustainability.

Weaknesses of the reviewed approaches for characterizing sustainability can be

related to their failure to incorporate the proposed elements (Table 2-6). Characterization

based on adherence to prescribed approaches fails because it is not founded on a literal

interpretation of sustainability. Lack of integration limits the usefulness of multiple

indicators of sustainability for diagnosing and prioritizing constraints. Integration of

indicators has been difficult because the underlying interpretation of sustainability as an

ability to meet diverse goals is not integrated. A time trend represents an integrated

system response that is potentially useful for diagnosis and can be predictive by

extrapolation, but it is not stochastic in the sense of accounting for variability. An

integrated measure of resilience has not yet been found. The assumptions about future

variability and disturbances necessary for resilience to be predictive and stochastic are

avoided in discussions of its use for characterizing sustainability. Simulated farm

survivability is the only approach reviewed that incorporates all of the elements listed.















Table 2-6. Approaches to characterizing agricultural sustainability. A check (V) indicates the approach incorporated the specified
element. A question mark (?) indicates it addressed some aspect of the element.

Approach Reerence Literal System- Quanti- Predic- Sto- Diag-
oriented tative tive chastic nostic

Reduced use of chemicals relative to other farmers in South Dakota Dobbs et at., 1991

Quantitative index of cabbage farmer practices in Malaysia Taylor et al., 1993 7

Adoption of alternative practices and reduced chemical use on U.S. Pacific Northwest farms Cordray et al., 1993 /

Indicators of resource base, system performance &external effects in agroforestry gardens Torquebiau, 1992 / /

Indicators of visible, masked, & potential degradation in Hymalaian farming systems Jodha, 1990 / 7

Indicators of regional "agroecosystem health" Neher, 1992 ?

Index of assured output, and output per unit input and per unit limiting resource decline Lal, 1991 / /

Index of productivity, stability and degradivity Sands & Podmore, 1993 / /

Subjectively weighted index of constraints to nine system attributes Stockle et al., 1994/ /

Non-negative time trend in output Lynam & Herdt, 1989 / / / 7

Total factor productivity accounting for natu-ral resources in cropping systems in Nigeria Ehui & Spencer, 1992 / / /

Regional trends in inputs and outputs in India Monteith, 1990 / / / 7

Properties representing resilience Conway, 1994 / / ? ?

Production trends and resilience in shifting cultivation in Malaysia Cramb, 1993 / / 7

Simulated trends and variability in crop production and NO; leaching Singh & Thornton, 1992 / / / / /

Simulated crop production trends in response to simulated soil erosion in Canada Lerohl, 1991 / / / /

Simulated survivability of Texas rice farms Perry etal., 1985 / / / / / /













Conclusions



The importance and desirability of agricultural sustainability are generally

recognized. However, its potential as a criterion for guiding agriculture's response to

change has not been realized. Characterization is a prerequisite to using sustainability as a

basis for guiding change. Logical inconsistencies limit the usefulness of characterization

of sustainability interpreted as an ideological or management approach to agriculture.

Interpreting sustainability as an ability to meet diverse goals suggests measuring sets of

system indicators consistent with those goals. However, these measurements have proven

difficult to integrate and interpret in a way that identifies constraints or focuses research.

Literal interpretations of sustainability as an ability to continue into the future suggest

measurable, integrated criteria for its characterization. However, applications of these

criteria--time trends and resilience--have ignored or misinterpreted important aspects of

system behavior. Criteria are needed that relate levels, trends and variability of long-term

system performance to the needs and goals of farmers and of society.

In order for sustainability to be a useful criterion for guiding change in agriculture,

its characterization should be literal, system-oriented, quantitative, predictive, stochastic

and diagnostic. These elements identify weaknesses in existing and proposed approaches,

suggest directions for future development of approaches, and together constitute a

systems approach for characterizing sustainability of agricultural systems. The tools of

system analysis and simulation must be part of approaches that incorporate these elements.














CHAPTER 3
A SYSTEMS FRAMEWORK FOR CHARACTERIZING FARM SUSTAINABILITY



Introduction



The potential benefits of applying the concept of agricultural sustainability--

providing feedback about future impacts of current decisions, and focusing research and

intervention by identifying constraints--can only be realized when sustainability of

particular systems is characterized. Characterization includes both quantification and

diagnosis of constraints.

In spite of the tremendous amount of concern about agricultural sustainability,

surprisingly few studies have attempted to characterize the sustainability of specific

agricultural systems. The methods which have been proposed or applied suffer from (a)

conceptual problems associated with interpreting sustainability as an approach rather than

a property of agriculture, and (b) practical difficulties that arise from the fact that

sustainability deals with the future (Chapter 2). Characterization based on management

practices either does not relate to a literal interpretation of sustainability or it leads to

circular logic. Attempts to characterize sustainability based on system response have

generally ignored or misinterpreted important system properties.














CHAPTER 3
A SYSTEMS FRAMEWORK FOR CHARACTERIZING FARM SUSTAINABILITY



Introduction



The potential benefits of applying the concept of agricultural sustainability--

providing feedback about future impacts of current decisions, and focusing research and

intervention by identifying constraints--can only be realized when sustainability of

particular systems is characterized. Characterization includes both quantification and

diagnosis of constraints.

In spite of the tremendous amount of concern about agricultural sustainability,

surprisingly few studies have attempted to characterize the sustainability of specific

agricultural systems. The methods which have been proposed or applied suffer from (a)

conceptual problems associated with interpreting sustainability as an approach rather than

a property of agriculture, and (b) practical difficulties that arise from the fact that

sustainability deals with the future (Chapter 2). Characterization based on management

practices either does not relate to a literal interpretation of sustainability or it leads to

circular logic. Attempts to characterize sustainability based on system response have

generally ignored or misinterpreted important system properties.












Since sustainability deals with the future, it cannot be readily observed. Analysis

and simulation of a model system can compensate for the limitations of observation and

experimentation on a real agricultural system. In this chapter, I first present a definition of

sustainability that applies generally to dynamic, hierarchical, stochastic, purposeful

systems. I then describe a framework for using system simulation to quantify sustainability

and test hypotheses about its constraints. A Monte-Carlo study of a simple time-series

model demonstrates how sustainability relates important components of system behavior--

mean, trend, variability and autocorrelation--to threshold goal levels. I then discuss issues

that arise when applying the framework to farming systems. Finally, I illustrate an

application of the framework for characterizing farm sustainability using data from a

previously published farm simulation study.



Defining Sustainability



To sustain is literally "to keep in existence; keep up; maintain or prolong"

(Neufeldt, 1988, p. 1349). Sustainability can therefore be defined as the ability of a

system to continue into thefiuture. Key words of this definition suggest a framework for

quantification.

First, the system addresses the question, "What is to be sustained?" Although the

system of interest in this paper is a farm, the concept of sustainability can be applied to any

system that is dynamic, stochastic, and purposeful. The ability to continue does not apply

to a static system. Furthermore, sustainability has no meaning unless some purpose or










35

threshold condition exists which distinguishes a system that is sustaining from one that has

failed. Finally, sustainability of a deterministic system is binary; in the absence of

uncertainty one can say that a system will either sustain itself or fail during some future

period.

Second, the word continue implies the possibility that a system can fail if some

criteria are met. Failure occurs when a system can no longer fulfill its purpose. Failure

implies irreversibility: a degree of stress from which the system cannot readily recover.

Criteria for failure address the question, "Above what minimum level is the system to be

sustained?"

Third, the future suggests a time period which extends from the present (t = 0) to

some future time (t = 7). For a stochastic system, the future also implies uncertainty;

uncertainty distinguishes the future from the past. The time period addresses the question,

"How long is the system to be sustained?"

Finally, the ability of the system to continue in the future is best expressed as a

probability. The suggestion that "A farm is either sustainable or it's not sustainable ... A

farm will either last for a very long period or it won't" (Rodale, 1990, p. 273) expresses a

common deterministic interpretation of sustainability. However, one cannot determine

with certainty whether a system will continue through some future period. Probability of

continuation provides a measure of sustainability with a zero-to-one range that addresses

the question, "With what degree of certainty will the system sustain itself?" The definition

of sustainability can be restated as the probability that a particular system will not meet

specified criteria for failure during a particular future period.













Quantifying Sustainability



Consider the status of a system, D(t), as a Bernoulli process with state space {0,

1} operating in the period from the present (t = 0) to some future time (t = T) (See symbol

definitions in Table 3-1). Continuation is indicated by D(t) = 1 and failure by D(t) = 0.

The system is initially operating: D(0) = 1. Time of failure, TF, is then a random variable

with a probability density function, f,(t) = P{TF = t), and a cumulative distribution, FrT(t)

= P{Tp t}. The distribution FTF applies to the population of possible time paths, or

realizations, of system behavior. Failure is irreversible such that if D(t) = 0 then D(t+At) =

0 for all At > 0. For the period (0, T], sustainability, S, is defined as,


S(T) = 1 FTr(T). [3-1]


The definition given in Eq. [3-1] is equivalent to the survival function in mortality studies

(eg., Elandt-Johnson and Johnson, 1980) and to reliability in quality control literature (eg.,

Barlow, Proschan and Hunter, 1965).


Failure as violation of state thresholds


Time to failure, T,, is a random variable because the system's state, x(t), behaves

as a stochastic process as it responds to a stochastic environment. Consider a single,

continuous state variable, x(t). At any given time t, x(t) has a probability density function,

fx,,(x), and a cumulative distribution, Fx,(x) that apply to an initial population selected at













Quantifying Sustainability



Consider the status of a system, D(t), as a Bernoulli process with state space {0,

1} operating in the period from the present (t = 0) to some future time (t = T) (See symbol

definitions in Table 3-1). Continuation is indicated by D(t) = 1 and failure by D(t) = 0.

The system is initially operating: D(0) = 1. Time of failure, TF, is then a random variable

with a probability density function, f,(t) = P{TF = t), and a cumulative distribution, FrT(t)

= P{Tp t}. The distribution FTF applies to the population of possible time paths, or

realizations, of system behavior. Failure is irreversible such that if D(t) = 0 then D(t+At) =

0 for all At > 0. For the period (0, T], sustainability, S, is defined as,


S(T) = 1 FTr(T). [3-1]


The definition given in Eq. [3-1] is equivalent to the survival function in mortality studies

(eg., Elandt-Johnson and Johnson, 1980) and to reliability in quality control literature (eg.,

Barlow, Proschan and Hunter, 1965).


Failure as violation of state thresholds


Time to failure, T,, is a random variable because the system's state, x(t), behaves

as a stochastic process as it responds to a stochastic environment. Consider a single,

continuous state variable, x(t). At any given time t, x(t) has a probability density function,

fx,,(x), and a cumulative distribution, Fx,(x) that apply to an initial population selected at












Table 3-1. Description of symbols used.
Symbol Description
t, T Time variable and a particular time
D(t) Status of a system continuingi, 0=failed)
TF Time to system failure

fT, F, Density and cumulative distribution of TF
S(T) Sustainability for the period (0, 7]
h(t) Sustainability hazard probability function
x, x State vector and a particular state variable

x0, xo Failure threshold values for x and x

fx,,, Fx,, Density and cumulative distribution ofx at time t
N Total number of realizations simulated
n(T) Number of realizations continuing at time T
S estimated from a finite number of realizations
SEs Standard error of S

z(t), oz A stationary stochastic process and its standard deviation
a, p Intercept and slope of a deterministic trend

C1 One period lag autocorrelation coefficient
e(t), o0 A white-noise process and its standard deviation

Yo, Y, Base and alternate value of the ith hypothesized determinant of sustainability
9o, ~i Sin response to the base and alternate value of the ith factor
Ri, ri Absolute and relative sensitivity to the ith factor
vab, N Frequency of a and b occurring, and sum of all frequencies
q Correction factor for continuity in G-tests

G1, Gadj Uncorrected and corrected statistic for tests of independent frequencies,
independent observations

Gp, Gp,dj Uncorrected and corrected statistic for tests of independent frequencies,
paired observations










38

time 0. Assume that a threshold value, x0, exists such that the system will fail the first time

x < xo. Ifxo represents a maximum threshold, a sign change is required to allow the

relationship to hold. We can express the relationship between x(t) and TF by considering

the probability that failure occurs during an interval (t, t+At], At > 0:


P{t< TF r t+At} = P{TF t+At I T> t} P{TF> t}, [3-2]


and, since the condition T, < t+At is equivalent to x(t+At) < xo, we can rewrite Eq. [3-2]

as,


P{t< T < t +At) = P{x(t +At) sxo I x(t)>xo (1 -P{TF t}). [3-3]


As At approaches 0, the left side of Eq. [3-3] becomes the probability density of time to

failure:


lim
tlm P{t< T,< t+At = P{T,=t}

= fTF() [3-4]



Since Fx,, is the distribution ofx for the population of realizations that have continued to

time t, Fx,t(xo) expresses the probability that the system violates the threshold xo at t, given

that it has not done so previously. Therefore, taking the limit of the first term on the right

side of Eq. [3-3] gives,


lim
lt 10 P{x(t +At) < xo I x(t) > x0} = Fx,(xo). [3-5]
Atl 0










39

Finally, the last term in Eq. [3-3] is equivalent to sustainability from Eq. [3-1]. By

taking the limit ofEq. [3-3] as At approaches 0 and substituting the simplified terms (Eq.

[3-4] and [3-5]), we obtain,


fTF(t) = Fx,t(xo) S(t). [3-6]


We can now derive an expression of S as a function of only Fx,,. Differentiating Eq. [3-1]

gives,

dS(t) / dt = -fT,(t). [3-7]


Substituting Eq. [3-6] into [3-7] gives the differential equation,


dS(t) / dt = Fxt(x) S(t), [3-8]


which has the solution,


S(T) = exp( fTFx,(Xo) dt [3-9]



at time T. Thus, I have shown that sustainability is determined entirely by the probability

that a system's state falls below a threshold value during some time interval (0, T].

Two examples illustrate how time, thresholds, and the distribution of system state

interact to determine sustainability. In the first example (Fig. 3-la), the variability ofx is

relatively high but its expected value, E[x(t)], remains constant. The system has a 0.15

probability of failing in any particular period. Sustainability declines exponentially with























period, t


period, t


S0.8 O8
0.8
0.6 0.6

c: 0.4 M0
- C' 0.4
I 0.2 cn
n :3 0.2
0.0 I I I I I 0.0 ,
0 1 2 3 4 5 6 0 1 2 3 4 5 6
period, t period, t

Figure 3-1. Relationship between time, distribution of state, and sustainability under (a) constant mean and high variability,
and (b) negative trend and low variability.












time as Eq. [3-9] predicts. In the second example (Fig. 3-1b), variability is lower, but

E[x(t)] decreases linearly with time. Here, the effect of the negative trend in x on

sustainability is not apparent until E[x(t)] approaches x0.

Several generalizations can be applied to the preceding discussion. First, a failure

threshold may exist for more than one system state variable. Second, the thresholds may

be dynamic. The state of a complex system and the corresponding set of minimum

threshold values could then be expressed as vector processes: x(t) and x0(t). The

threshold vector xo(t) bounds the system's state space in one or more dimension. Since

only a few of the state variables may be directly related to the ability of a system to

continue, many members of x(t) may have values of -o for all t. Third, the threshold

vector could be stochastic. The probability of violating system thresholds would then

depend not only on the behavior of x(t) but also on the dispersion of xo(t) and its

correlation with x(t). Finally, thresholds may apply to derived state variables such as a

sum (eg., aggregate wealth) or ratio (eg., a financial ratio) of basic state variables.


Sustainability hazard


The value, Fx,,(xo), is the instantaneous probability of failure applied to the

population of realizations that have continued up to time t. It is referred to as a hazard

function,


h(t) = fTrt) / (1 FT(t))

= f(t) / S) [3-10]














(Barlow et al., 1965). Hazard is a probabilistic expression of the intensity of stress on a

system as a function of time; increasing h indicates an increasing stress or increasing threat

to sustainability. By rearranging and substituting Eq. [3-10] into [3-7], and integrating,

we arrive at,


S(T) = exp h(t) dt. [3-11]




We see that constant hazard results in exponentially declining sustainability. Although S(t)

increases monotonically from an arbitrary starting time (t = 0), the threat to sustainability

expressed by h(t) is independent of starting time, and may increase or decrease.



Simulating Sustainability



Equations [3-1] and [3-9] cannot be used to calculate sustainability in practice

because the distributions of T, and x(t) are generally not known for a real system.

Harrington et al. (1990) expressed the need to artificially construct time paths for current

and alternative strategies to assess sustainability before long-term experiments or

monitoring could be completed. A set of replicated time paths, or realizations, can be

sampled by stochastic simulation of a model of the system.

Consider a set of N realizations simulated for the period (0, T] (Fig 3-2). Although

the environment is sampled randomly, each replicate has the same initial conditions, x(0).









43







S- -* -x2 )W



Z to-) wo
/ .-"^^ 4% -



.. ,xW(t)




xI------------.x--s----






1.0




S(T) ------------------

C


=3
4 4.










4~ 4---------- 4-+








0.0

To TF3 TF5 T

time, t


Figure 3-2. Estimating sustainability by sampling a small number (i.e. 5) of simulated
realizations of future system behavior.
",x()_ '._5t








*1n
Co
c (T












Let n(t) be the number of realizations continuing at time t. Sustainability can then be

estimated by the relative frequency of surviving realizations, or


S(T) = n(T)IN. [3-12]


Because of the uncertainty associated with estimating the probability of success from a

small Bernoulli trial, sustainability estimated from Eq. [3-12] has a standard error

(Snedecor and Cochran, 1980) of


SEg = /(T) (1 S(7)) /N

= n(T)(N- n(T)) N3. [3-13]



An example: sustainability of a simple time-series model


The proposed definition (Eq. [3-1]) integrates several important aspects of system

behavior--means, trends, variability and autocorrelation--and relates them to levels of

goals expressed as system thresholds. I used Monte Carlo simulation of a simple, discrete-

time, univariate time series model to demonstrate the role of these aspects of system

behavior in determining sustainability. The model consisted of a deterministic trend

component,


x(t) = a + pt + z(),












and a first-order autoregressive component,

z(t) = z(t- 1) + e(t),

where a and p are intercepts and slope of a time trend, z(t) is the value of a stationary time

series in period t, 4( is a first-order autoregressive coefficient, I~4 | < 1, and N(0, ao) is

a random shock variable. The standard deviation of the random shock (o() is related to

the asymptotic standard deviation of the generated time series (oz) by,



Gc = _, .



(Pankratz, 1983). I calculated S by Eq. [3-12] from 10,000 replicated realizations, each

simulated for 120 periods.

Table 3-2 summarizes values of the parameters and the results of the sensitivity

analysis. Sustainability, S(120), was 0.567 0.005 ( S.E.) in response to the base

parameters. Decreasing the mean, increasing the goal threshold, or increasing variability

reduced 5(120) by increasing the proportion of the population which fell below x0 at any

time before t=T. Decreasing autocorrelation reduced ( 120) because of its effect on Fx,,.

For discrete time periods, Fx, in any period I is conditioned on the prior status of the

system and, therefore, on the previous value of the state variable. Finally, extending the

time period Tbeyond 120 decreased S(T).

Figure 3-3 shows how declining expected value, and abrupt changes of expected

value, variance and autocorrelation influence the sustainability and hazard time functions













of the time-series model described above. The general pattern of the sustainability time

function remains the same whether hazard is modified by a change in the expected value,

variance or autocorrelation (Fig. 3-3c-h).



Table 3-2. Sensitivity of simulated sustainability to system properties. !120)=0.567 for
the base scenario.


Property

mean

trend

variability

autocorrelation

goal threshold

duration


Base
value

10.0

10.0
0.0

5.0

0.8

3.0

120


------Increased------
value ST)

11.0 0.783

9.0
0.566
+0.0167

5.5 0.401

0.88 0.924

3.3 0.491

132 0.535


-----Decreased------
value S(T)

9.0 0.307

11.0
0.517
-0.0167

4.5 0.774

0.72 0.235

2.7 0.641

108 0.604


Diagnosing Constraints to Sustainability


The potential value of the concept of sustainability lies in its ability to focus

research and intervention by identifying and ranking its constraints. Diagnosing

constraints entails a process of hypothesis formulation and testing using simulation of the

system model. Hypotheses should identify the current (or expected) value of a suspected

constraint, and a specific change that would relax the constraint. Sensitivity analysis then

provides the experimental tool for testing and ranking hypothesized constraints.


Para-
meter

a

a


(3,


xo
T















0.8. 60
12 /
.60.6 10 40
8
0..4 6
4 20
<( 0.2 / P = -0.045 \
.2 o,2= 4.0 B
0.0 I 0 I 0

1.0 25
C D
.0.8 20 30
oI 20

6 15 20

S0.4 I
10 1
0.2 5 5
a = 8.0 a =10.0 /a =10.0 a 8.0
0.0 0 0
1.0 16 16
E 14 F 14
0.8
12 12
.G0.6 10 10
S 10
.0.4 / 16 6

0.2 / N
So =9.0 =7.0 7.0 = 9.0
0.0 0 O0
1.0
r\ ^ -^ H ^ 20 E
,0.8 15

i0 0.6
/ 10
S / 10
C 0.4 -


= 0.87 $ =0.93 =0.87
0.0 0 0
0 40 80 120 0 40 80 120
Time, t Time, t


Figure 3-3. Sustainability and hazard of an AR(1) process with (a) constant parameters,
(b) declining expected value, (c) abrupt decrease and (d) increase in mean, (e) decrease
and (f) increase in variance, and (g) decrease and (h) increase in autocorrelation.
Parameter values were a = 10.0, 13 = 0.0, oa = 8.0, 4j = 0.9, x0 = 4.0 and n = 50,000
except as indicated otherwise.











Sensitivity analysis


Sensitivity analysis is used to quantify the relative importance of hypothesized

constraints to sustainability. It involves changing the value of a factor a small amount in

the direction that would relax the hypothesized constraint relative to a base scenario which

represents existing or expected conditions, then simulating the modified scenario.

Hypothesized constraints can then be ranked based on either absolute,


Ri= Si S [3-14]


or relative sensitivity,


Yi, -o
ri [3-15]



where Y,o is the value of the ith factor in the base scenario, Y, is its adjusted value, and 90

and are sustainability values estimated for the base and alternate scenarios. The

absolute value allows an increase in sustainability to result in ri > 0 regardless of the

direction of change in Yi. Relative sensitivity is interpreted as the percent change in g in

response to a 1% change in Y. Comparisons may be made and ranks assigned among

discrete or among continuous factors. However, absolute sensitivity to discrete factors

cannot be compared with relative sensitivity to continuous factors.












Significance tests


Independent observations. The frequencies (vij) of failure and continuation in the

simulation of the base and alternate scenarios used for a hypothesis test can be represented

by a two-way contingency table arranged as in Table 3-3, where i is the row andj is the

column. Several tests are available for the null hypothesis that frequencies are

independent, equivalent to the null hypothesis that sustainability is the same in the two

scenarios. Sokal and Rohlf (1981) recommended a log likelihood ratio, or G-test, over

the more frequently used X2 test. One can calculate the expected frequencies of

continuation or failure from the observed frequencies based on the null hypothesis of

independence:

j = (Vi1 Vi2) (VIj + j) /N

The GI statistic is then calculated from the observed and expected frequencies:

2 2
G, = 2 E E(vij n(vij /i)), [3-16]
i=1 j=1



and is corrected for continuity (Williams, 1976):


q = 1 + ((N/v1 + N/v2, 1)(N/v1 +N/v*2 1))/6N, [3-17]


Gadj =G q.


[3-18]












The G statistic is distributed approximately X2 with one degree of freedom. If Gldj > X 2

then reject Ho: S(T) = Si().




Table 3-3. 2x2 frequency table for tests of difference between simulated
sustainabilities, independent observations.
Scenario continued failed Total
base Vn V12 V* = V1+V12
alternate V21 V22 V2 = Y21 V22
Total: vi*=V11+v21 v2=V12+V22 N = ij




Paired observations. Different scenarios can be replicated under the same set of

environments by using the same pseudorandom number sequence for sampling stochastic

inputs. The additional information available from such a randomized block design permits

the use of a more powerful test. When data are arranged as in Table 3-4, the row or

column totals then represent the frequency of failure or continuation of the base (row

totals) or alternate (column totals) scenarios. Then vn is the number of replicated

environments in which both scenarios failed, v12 is the number in which the base scenario

failed but the alternate scenario continued, and so forth. The McNemar (1947) test as

adapted by Sokal and Rohlf(1981) uses a G statistic calculated as,












Gp=2 v In 12 V21 + v,, In 2V21 [3-19]



with the correction factor,

q = 1 + 1/2N, [3-20]


applied as in Eq. [3-18] (Williams, 1976). Again, G,adj is compared to X2 with one degree

of freedom. The McNemar test statistic will be undefined if any of the frequencies (Table

3-3) have a value of zero. This can easily occur if, for example, none of the replicates that

continue in a base scenario fail in an alternate scenario (v12 = 0).




Table 3-4. 2x2 frequency table for tests of difference between simulated
sustainabilities, paired observations.
Base --Alternate scenario--
scenario continued failed Total
continued V1n V12 V1* = n 12
failed v2 v22 v2* = v21+v22
Total: v*, = vu+v21 v*2 = v12+ N = v




It is important to keep in mind that statistical inferences based on simulated

sustainability apply to the model system and its environment. Extension to the actual

system depends on the validity of the model and assumptions about the future

environment.













Sustainability Applied to Farming Systems



Although sustainability is an important concern at several levels in the hierarchy of

agricultural systems (Lowrance et al., 1986; Lynam & Herdt, 1989), it is particularly

relevant at the farm level. If agriculture is to meet the needs of society--providing food

and other products while protecting natural resources--it must first meet the needs of the

farmers who implement and manage it.

The framework for characterizing sustainability applies to farming systems because

a farm is a dynamic, stochastic and purposeful system. Furthermore, sustainability is

characterized most easily at the farm level where system goals are more easily specified

and more consistent than at other system levels. For example, human goals are not

intrinsic to fields or enterprises. On the other hand, the emergence of many human actors

at levels higher than the farm leads to multiple and often conflicting goals. The continuing

and failed status would comprise fuzzy, nonexclusive sets at these higher system levels.

Sustainability expressed as a probability of continuation (Eq. [3-1]) must be

applied to some initial population. One could think of a population of farms and attempt

to predict the proportion that will survive through a future period. However, the

appropriate population to consider when characterizing farm sustainability is the

population of possible realizations of future behavior of an individual farm.

A farm's context within hierarchy has implications for selecting an appropriate

time frame for sustainability analysis, identifying failure criteria, making assumptions about










53

the future behavior of system inputs, and hypothesizing constraints. The remainder of this

section examines these issues.


Selecting a time frame


Sustainability has meaning only in the context of a specific time frame. For

example, climate change, soil erosion, or extinction may be seen as irreversible threats to

the sustainability of an ecosystem in a time frame of decades or centuries, but as part of

natural cycles in a time frame of millennia or longer (Fresco and Kroonenberg, 1992).

From another perspective, sustainability can be viewed as a non-increasing function of

time. Considering the extreme cases, all existing agricultural systems can sustain

themselves for an arbitrarily short period. On the other hand, few agricultural systems can

be expected to continue in a recognizable form for tens of millennia. IfFT(t) is a true

probability distribution with a lower bound at t=0 then


lim
1-0o FTF(t) = 1.



Although selecting a time period for analyzing farm sustainability is a subjective

decision, considerations of hierarchy, relevance, and realism suggest a range of about 10

to 15 years. Ecological hierarchy theory states that processes in higher-level systems

operate more slowly than in lower-level systems (Allen and Starr, 1988). The time frame

for analyzing sustainability of a farming system should therefore be longer than the several

months to a few years that are typical of crop and animal production cycles. Relevance












suggests that the time interval should be long enough to allow detection of important

threats to sustainability. For example, three or four years would reveal little about the

impact of soil erosion on sustainability. On the other hand, a study exceeding a century

would not be relevant to the livelihood goals of an individual farm. Lynam and Herdt

(1989) suggested that five to 20 years is a relevant time-frame for analysis of farming

system sustainability. Realism of assumptions about economic, policy, and technological

inputs to the farming system becomes increasingly difficult to defend past about 10 to 15

years.


Assumptions about inputs


The higher-level systems that comprise a farming system's environment exert

control through inputs and through constraints to farmer decisions or farm outputs.

Analysis of farm sustainability requires assumptions about future behavior of inputs and

control mechanisms that are conceptually external to the farming system. The issues to

consider include (a) whether systematic trends or cycles are expected, (b) whether

variability is sufficient to warrant stochastic sampling, and (c) whether a feedback

mechanism exists which allows an input to respond to farm outputs. Although some farm

processes may be inherently stochastic, inputs of weather and prices are usually the major

sources of risk in a farming system. Catastrophic events such as disease epidemics, storms

or wars also represent important stochastic inputs to some farming systems.












Farm failure criteria


Failure criteria denote the minimum level of performance above which a system is

to be sustained. Since farmer livelihood is the primary purpose of most farming systems,

criteria for farm failure can be expressed in terms of minimum levels of livelihood goals.

Hamblin (1992) suggested that agriculture fails to sustain if production falls below the

levels necessary for profitability in a cash economy or survival in a subsistence economy.

In a subsistence economy, a level of poverty or malnutrition from which a farm family

cannot escape without outside intervention might indicate system failure. Lynam and

Herdt (1989) referred to famine as "the ultimate indicator ofunsustainability" (p. 391). In

a cash economy, lenders may impose threshold leverage ratios above which they will force

foreclosure by recalling loans (Perry et al., 1985). Failure could be expressed in several

forms such as farm abandonment, conversion of land to non-agricultural use, the need to

supplement income with off-farm employment, inability to meet critical goals such as

education of children, or major changes in farm enterprises, depending on the analyst's

purpose.

Negative feedback loops between components of a system tend to counteract the

effects of disturbances and stabilize a system, resulting in a stable state or attractor.

Mathematical and empirical evidence suggests that ecosystems can possess multiple stable

states (May, 1977). The region about a stable state in state space is a domain of

attraction, and the boundary between adjacent domains of attraction is a separatix

(Trenbath et al., 1990). If the state of a system is displaced across a separatix, it enters a












new domain of attraction. The ability to return to the original domain of attraction

depends on the relative stability of the two domains, the nature of the separatix between

them, and the existence of disturbances which could displace the system back across the

separatix. Failure of an agricultural system can be viewed as transition from a useful to a

less useful domain of attraction. The state threshold vector x0 forms the separatix between

the domains. Trenbath et al. (1990) used mathematical models to illustrate abrupt

transitions from useful to less useful domains in response to intensification of three

agricultural systems.

In many cases, farm failure criteria may be difficult to determine. However, it may

be possible to obtain meaningful insights into the relative impact of various stresses on a

farming system by assuming particular failure thresholds when those thresholds cannot be

measured.

Since increasingly restrictive failure thresholds increase the probability of system

failure, sustainability may be viewed as a non-increasing function of threshold levels. A

system is less able to continue at a high level than at a low level.


Determinants of farm sustainability


Sustainability is an aggregate response of a system to a range of external factors,

conditioned by internal characteristics of the system. Any factor that influences means,

trends, variability, autocorrelation or goal levels may influence sustainability. A host of

factors influences the balance between income and expenditure that determines the mean

level of farm wealth. Soil degradation, depletion of scarce resources, technological












innovation and trends in prices can affect trends in farm state variables. Variability is

influenced by weather patterns, price volatility and the occurrence of catastrophic events.

Credit availability, market access, and the ability to store agricultural products have

positive effects on autocorrelation of farm wealth. Finally, a household's tolerance to

difficulty, alternative sources of livelihood and lenders' policies can influence goal

thresholds.

Adaptive management generally serves to improve sustainability. Farmers employ

a range of management strategies, such as selling capital assets, reducing input use,

working off-farm, or shifting from cash to subsistence crops, to reduce the risk of failure

during difficult times. Although the proposed framework for characterizing sustainability

can account for adaptive strategies, the farmer decision process may be more difficult to

simulate than biological or economic processes. Simulating a fixed management strategy

may greatly overestimate the probability of farm failure.

Much of the concern about sustainability of agricultural production systems relates

to externalities: non-target outputs with costs (or benefits) which are not born by an

individual farm but by society. A hierarchical perspective suggests that externalities

should have no direct impact on farm sustainability; the impact of an externality emerges

at a higher (eg., regional) system level. However, a feedback mechanism that either

charges the farming system for the externality or constrains the practice that produces it

can easily be incorporated into a simulation analysis. Such a deviation from a strict

hierarchical approach might be justified by assuming that society will attempt to correct

the perceived injustice of a negative externality.










58

A similar issue arises when extending an analysis from a single farm to a group of

similar farms. A group of farmers could adversely affect common grazing land, water

resources, fuelwood, fishing areas, or forest by overuse, whereas a single farmer's impact

might be negligible. Similarly, although an individual farmer is usually assumed to be a

price taker, a group of farmers may alter prices due to their aggregate effect on supply of

a product or demand for an input such as seasonal labor. Incorporating these effects

involves extending the analysis to include some processes (eg., market equilibria) above

the farming system level.



An Example: Sustainability of a Coastal Texas Rice Farm



Whole-farm simulation studies have examined the influence of factors such as

commodity price variability (Grant et al., 1984), farm size and beginning equity level

(Richardson and Condra, 1981), intergenerational estate transfer strategy (Walker et al.,

1979), and land tenure expansion strategy (Held and Helmers, 1981) on probability of

farm survival through various periods. Perry et al. (1986) conducted a more

comprehensive simulation study that examined the impacts of crop rotation, land tenure

arrangement, government programs, costs, labor availability, lenders' policies, interest

rates, and the level and variability of crop yields and prices on rice farms in Texas. A

reinterpretation of the results of this study illustrates the use of long-term, stochastic

simulation to characterize farm sustainability.












Methods


Perry et al. studied a "representative" coastal Texas rice farm rather than an actual

farm. They examined four combinations of rotation and sharecropping arrangement.

Assumptions and initial conditions can be found in Perry et al. (1986).

Perry et al. used a modified version of the FLIPSIM (Firm Level Income Tax and

Farm Policy Simulator) simulation model (Richardson & Nixon, 1985) called RICESIM to

simulate the model farm. RICESIM is primarily a farm accounting model which randomly

samples from probability distributions as a proxy for the biological and ecological

processes involved in crop production.

The study examined two alternate criteria for system failure. First, probability of

survival was based on a threshold leverage ratio (total debt/total equity). Lenders were

assumed to force foreclosure by recalling loans if the leverage ratio exceeded 2.0. The

second criterion, a negative net present value (NPV) of future cash flow, requires a higher

level of minimum system performance to avoid failure. Failure indicated by a negative

NPV means that a secure, non-farm investment would be more profitable than farming.

Probabilities of survival and positive NPV were calculated by Eq. [3-12] from 50

replicates of each five-year scenario.

Perry et al. analyzed sensitivity of the probabilities of survival and positive NPV to

several factors. For our analysis, I selected only the soybean-soybean-rice rotation with

1/2 share of rice and 1/7 share of soybean going to the landowner. I tested those factors

included in the sensitivity analysis that I believed could constrain sustainability (Tables 3-5












and 3-6). To test for differences in sustainabilities, I used a G-test that is based on

independent sampling (Eq. [3-16] to [3-18]) since the study did not provide paired data or

indicate the experimental design.


Results


The values of Sat the end of the five-year scenario presented here were 0.50

0.071 based on leverage ratio and 0.12 0.046 ( S.E.) based on NPV. Figure 3-4 shows

S(t) of all four combinations of crop rotation and tenure arrangement based on the

threshold leverage ratio.




Table 3-5. Relative sensitivity, r (Eq. [3-15]), of simulated five-year sustainability, S5), of
a Texas rice farm to continuous factors. For the base scenario, (5) was 0.50 based on
threshold leverage and 0.12 based on threshold NPV.
Base Alternate Leverage < 2.0 ---- NPV > 0.0 ----
Factor value value S(5) r S(5) r
variable costs 100% 90% 0.82 6.40 ** 0.48 30.00 **
crop share 50% 45% 0.80 6.00 ** 0.46 28.33 **
crop pricest 100% 110% 0.68 3.60 n.s. 0.38 21.67 **
mean rice yield 100% 110% 0.60 2.00 n.s. 0.26 11.67 n.s.
mean soybean yieldt 100% 110% 0.58 1.60 n.s. 0.26 11.67 n.s.
ratoon red rice 25% 15% 0.80 1.50 n.s. 0.32 4.17 n.s.
rice yield variancet 100% 75% 0.54 0.32 n.s. 0.12 0.00 n.s.
soybean yield variancet 100% 75% 0.52 0.16 n.s. 0.10 -0.67 n.s.
tPercent of base value.










61

Table 3-6. Absolute sensitivity, R (Eq. [3-14]), of simulated five-year sustainability, S5),
of a Texas rice farm to discrete factors. For the base scenario, 9(5) was 0.50 based on
threshold leverage and 0.12 based on threshold NPV.
Base Alternate Leverage < 2.0 -- NPV > 0.0 --
Factor value value S(5) R S(5) R
tenure
tene1/2 share 100% owned 1.00 0.50 ** 0.32 0.20 *
arrangement

tenure
r ent 1/2 share 50% owned 1.00 0.50 ** 0.18 0.06 n.s.
arrangement

soybean
sirrion non-irrigated irrigated 0.86 0.36 ** 0.64 0.52 **
irrigation
tenure
a en 1/2 share 1/7 share 0.82 0.32 ** 0.52 0.40 **
arrangement

tenure rent at
1/2 share 0.74 0.24 0.56 0.44 **
arrangement $74/ha

rotation SSR SR 0.72 0.22 0.20 0.08 n.s.

loan interest 2 points
Si variable 0.54 0.04 n.s. 0.22 0.01 n.s.
rates lower

ratoon rice 7% quality no quality 0.50 0.00 n.s. 0.12 0.00 n.s.
quality discount discount


Of the continuous factors tested in the sensitivity analysis, only variable costs and

the rice crop share showed a significant role in constraining sustainability (Table 3-5). The

ranking of continuous factors by relative sensitivity was the same for sustainability based

on the threshold leverage ratio or on positive NPV. Changes in most of the discrete

factors showed a significant improvement in sustainability based on leverage ratios (Table

3-6). However, the ranking of discrete factors was different when sustainability was based

on NPV. These results suggest that reducing variable costs, negotiating more favorable











tenure arrangements, and irrigating the soybean crop would be the most important

strategies for enhancing sustainability of the model farm.



Discussion


By defining sustainability as the ability of a dynamic, stochastic, purposeful system

to continue into the future, I arrived at a useful, quantitative expression of sustainability.


1.0


0.9


>0.8

CU
cr
.%0.7
.6-
c'o

0.6


0.5


0.4
1984


1985 1986 1987 1988


Year

Figure 3-4. Simulated sustainability of a Texas rice farm under four scenarios: a
three year soybean-soybean-rice rotation with a 1/7 (SSR 1/7) and a 1/2 (SSR 1/2)
share arrangement, and a two year soybean-rice rotation with the same two share
arrangements (SR 1/7 and SR 1/2). Data from Perry et al., 1986.












Although sustainability of a real agricultural system cannot be observed because it deals

with the future, it can be estimated from simulation of a system model. Testing

hypotheses about constraints to system sustainability is then straightforward. Applying

this framework to farming systems results in an approach to characterizing sustainability

that is literal, system-oriented, quantitative, predictive, stochastic and diagnostic (Chapter

2). The use of such an approach could provide attempts to improve farm sustainability

with objective feedback.

The requirement for comprehensive and realistic farm simulation tools currently

limits application of the proposed approach. Most existing farm-level simulation models

are not sufficiently comprehensive; they do not integrate models of crop and animal

production, environmental degradation, economic processes, and farmer production and

consumption decisions. A study designed to examine a single constraint to farm

sustainability would be less demanding in its model and data requirements.

One could question the realism of assumptions about the future behavior of inputs

to a farming system that are required for characterizing its sustainability. However, all

approaches to characterizing sustainability involve inferences about the future. If high-

level systems change more slowly than lower-level systems, as ecological hierarchy theory

asserts (Allen & Starr, 1988), then future projections of inputs such as weather, prices,

infrastructure and technology are more defensible than extrapolation of past farming

system behavior.

The study by Perry et al. (1986) illustrates the utility of the framework presented

in this paper for characterizing sustainability. Although I interpreted their study beyond its











64

original purpose, their simulation analysis was comprehensive enough to illustrate how the

probability of continuation integrates the effects of a range of factors, and how sensitivity

analysis can be used to identify and rank those factors that constrain sustainability.














CHAPTER 4
AN OBJECT-ORIENTED REPRESENTATION OF A FARMING SYSTEM



Introduction



Chapter 3 presents a framework for applying simulation to characterize the

sustainability of a farming system. The framework calls for a farm simulation model that is

able to simulate the operation of a farm over an extended period and to replicate that

period with stochastic inputs of the important variables that contribute to farm risk,

particularly prices and weather.

A comprehensive review of existing farm simulation tools is beyond the scope of

this chapter. Fortunately, a few good reviews of farm modeling literature are available.

Klein and Narayanan (1992) provide an excellent summary of the history of farm modeling

efforts. Jones et al. (1995) derived several generalizations from a review of applications

of farm-scale models. Two are particularly relevant to this study. First, ".. there is

usually a new model for each study, with little acknowledgment of the potential for using

the same model for different farms by emphasizing data requirements and collection for

implementation" (p. 5). One of the exceptions is the FLIPSIM family of models

(Richardson & Nixon, 1985), which has been used to study the impacts of farm size

(Richardson & Condra, 1981), price variability (Grant et al., 1984), marketing strategies














CHAPTER 4
AN OBJECT-ORIENTED REPRESENTATION OF A FARMING SYSTEM



Introduction



Chapter 3 presents a framework for applying simulation to characterize the

sustainability of a farming system. The framework calls for a farm simulation model that is

able to simulate the operation of a farm over an extended period and to replicate that

period with stochastic inputs of the important variables that contribute to farm risk,

particularly prices and weather.

A comprehensive review of existing farm simulation tools is beyond the scope of

this chapter. Fortunately, a few good reviews of farm modeling literature are available.

Klein and Narayanan (1992) provide an excellent summary of the history of farm modeling

efforts. Jones et al. (1995) derived several generalizations from a review of applications

of farm-scale models. Two are particularly relevant to this study. First, ".. there is

usually a new model for each study, with little acknowledgment of the potential for using

the same model for different farms by emphasizing data requirements and collection for

implementation" (p. 5). One of the exceptions is the FLIPSIM family of models

(Richardson & Nixon, 1985), which has been used to study the impacts of farm size

(Richardson & Condra, 1981), price variability (Grant et al., 1984), marketing strategies












(von Bailey & Richardson, 1985), tenure arrangements and several other policy and

management factors (Perry et al., 1986) on farm viability. The first generalization

highlights the need for simulation tools and associated data standards that are generic,

flexible and extensible. The second generalization is that "... most farm-scale models

have been developed with a bias toward economics and limited consideration of the

biophysical components" (p. 5). Farm models often represent crop and animal response to

weather variability by sampling from probability distributions that were fitted to historical

or survey data. However, concerns about ecological threats to sustainability such as soil

degradation and climate change are better addressed by the integration of biophysical and

economic models. Recent farm models have incorporated crop simulation to capture the

impact of weather variability on farm risk (Dillon et al., 1989) and to relate global climate

change scenarios to farmer adaptation (Kaiser et al., 1993), for example.

A farm simulation model, the Farming System Simulator (FSS), was developed as

a tool for applying the framework presented in Chapter 3 for characterizing farm

sustainability. Its object-oriented design address the need for a generic simulator that can

represent very different farm types, and whose functionality can be extended relatively

easily. Linkage to process-level crop simulation models addresses the need for balanced

and integrated treatment of the ecological and economic components of a farming system.

The objectives of this chapter are (a) to describe the Farming System Simulator and (b) to

present data requirements for simulating the sustainability of a farming system. The

chapter first presents an overview of the design and functionality of FSS, then describes its

inputs, structure, processes and outputs.













Overview of the Farming System Simulator



The Farming System Simulator (FSS) is an object-oriented, dynamic, stochastic,

discrete-event farm simulator. It is object-oriented in its design, and is implemented using

the object-oriented extensions of Borland Pascal (Borland International, 1992). The

description of FSS that follows cannot be fully understood without some familiarity with

object-oriented programming concepts (Appendix A). FSS is dynamic; it is capable of

simulating the operation of a farming system through many years. It is stochastic in the

sense that it is designed to simulate many replicates of a farm scenario with stochastic

inputs of weather and price data, and to present analyses based on the resulting

distributions. FSS runs external crop simulation models to simulate continuous

physiological and ecosystem processes. However, all farm-level processes (i.e.,

operations, production, management and consumption of resources, and failure) occur in

response to discrete events. A final characteristic ofFSS is that it is primarily a resource

accounting model.

The object-oriented structure of FSS is based on a conceptual model of the

structure and function of a farming system. A farming system integrates ecological,

economic and social components (Fig 4-1). The ecological component of a farming

system--a set of agricultural ecosystems, or agroecosystems--consists of biotic

communities and the landscape that they inhabit, and can be delineated by field boundaries.

The economic component comprises the set of resources that are under the control of the












LOCAL
INDUSTRY?
I


S- I --
---------------------------------------


ecological
systems


economic social
systems systems


Figure 4-1. Parallel ecological, economic and social hierarchies of agricultural systems.


WATERSHED?
I-----;-----I


AGROECOSYSTEM
(LANDSCAPE &
BIOTIC COMMUNITY)


VILLAGE?
I ~ILG1


f












farmer. The social component of a farming system is the farmer or farm household, and

includes goals and decision criteria. In FSS, objects represent each of these components:

fields in the farm landscape, a set of strategies that specify the sequence of crop activities

and their management, a set of enterprises that link management strategies to particular

fields within the landscape, and a set of farm resources (Fig. 4-2). The household is

represented by decision rules for consumption, production and farm failure. The object-

oriented design of FSS provides a flexible means of representing farm resources, possible

interactions among resources, and relationships between operations and the resources that

they use. The Farming System Simulator is a farm model only in a loose sense; the model

structure of a particular farm is specified at run-time by the resource, field, enterprise and

strategy objects that are initialized in response to input data.

The capabilities of FSS reflect its purpose as a tool for characterizing farm

sustainability based on the framework presented in Chapter 3. The first requirement was

the ability to replicate a long-term scenario with constant initial conditions but stochastic

inputs. FSS takes advantage of a stochastic weather generator, WGEN (Richardson,

1985), that has been incorporated into the crop models that are part of the Decision

Support System for Agrotechnology Transfer version 3 (DSSAT, Hoogenboom et al.,

1994). An analogous stochastic price generator is part ofFSS. FSS addresses its second

requirement--an ability to simulate ecological processes of crop production--by calling

external crop simulation models. The crop models simulate weather variability, soil

dynamics and crop growth and development, then return the information that FSS requires

in the form of schedules of field operations. Third, FSS addresses the need to deal with












has a has a
set of has a set of


SResources


has a
set of
\


SProduction
decision criteria


Landscape
has a
set of
\





SGoal
thresholds


Enterprises

has a
set of
/
Fields has a
set of


Production
strategies


Figure 4-2. Object representation of the main components of a farming system.


has a


has a has a
set of set of
1


I Consumption
decision criteria












farmer livelihood by accounting for all farm resources produced, used for production, or

consumed by the farm household. The fourth requirement was the ability to test

conditions for system failure. Failure can be based on insolvency--the inability to cover

fixed costs, obligations or minimum subsistence consumption requirements. Failure can

also occur when an individual or aggregate farm resource violates a user-specified

threshold value. Detailed resource accounting was a prerequisite to the ability to test for

conditions for farm failure. Finally, FSS offers several file and graphical outputs that are

relevant to the analysis of farm sustainability.

The current version of FSS possesses several important limitations. First, it does

not simulate adaptive management; it simulates a fixed, continuously repeating set of

management practices. However, an actual farm operating under stress would normally

employ a range of practices to avoid failure (Chapter 3). Second, FSS does not possess a

mechanism for adjusting crop management for within-season resource constraints. This

limitation is imposed by the need to simulate each crop for an entire season. Chapter 5

discusses the problem and a possible solution. Third, FSS can simulate only crop

production enterprises. No livestock model has yet been adapted for running under FSS.

Fourth, the household resource consumption model (Eq. [4-7]) is simplistic; it does not

consider the impact of risk or anticipated future lifestyle changes on consumption and

savings decisions. The remaining sections of this chapter focus on inputs, processes and

outputs of FSS.













Inputs



Jones et al. (1995) cited a lack of emphasis on data standards as a barrier to

reusing farm models for different farms or applications. FSS input data structures and file

formats were designed to be flexible enough to be able to represent a range of farm types

and to accommodate possible future extensions of FSS. The general organization of data

and scheme for its use are being proposed as a starting point for developing data standards

for enterprise and farm-level systems analyses (Hansen et al., 1995). Detailed

presentation of input data requirements and formats in Appendix B supplements the

discussion in this section.

Afarm scenario is the operation of a farm through a period of time with a given

set of initial conditions and rules for making decisions and scheduling activities. A

scenario may be replicated with stochastic sampling of input variables such as weather and

prices, but with the same initial conditions and decision rules for each replicate. At least

two files--a scenario file and a price file--are required to simulate a farm scenario. The

scenario file contains farm-level information and identifies the other input files. The price

file contains the parameters for models for generating sequences of prices. A scenario that

calls external IBSNAT crop simulation models also requires a minimum data set (MDS)

for each crop consisting of a crop management file, soil file, weather or climate file, and

genetic coefficient file (Tsuji et al., 1994).












The set of input data required to run FSS serves three roles. First, inputs define

the structure of a farm model. The objects that define the model of a particular farm are

initialized from the input files. This is the topic of the next section. Second, inputs specify

initial conditions by specifying initial values of the state variables of objects. Third, inputs

provide the values of dynamic driving variables. The two types of driving variables--prices

and weather--can be represented either by inputs of actual historical series or as stochastic

time-series models.

The structure of the input files relates closely to the object-oriented structure of

FSS. In most cases, a line or section of an input file is an argument for the constructor

method (see Appendix A) that initializes an object. Multiple items within a section of a

file are used to initialize collections of similar objects. For example, each line in the

RESOURCES section of the farm scenario file is used to initialize and insert a resource

object into the collection of farm resources.


Scenario file


The farm scenario file contains the information necessary to initialize a farm model

and simulate a farm scenario. The scenario file also identifies other essential data files: a

crop management file, a soil file and a price file. Table 4-1 describes the sections of a farm

scenario file. Appendix B presents the format of each section.

Simulation control. The first three sections of the scenario file relate to simulation

control. The SCENARIO section identifies the farm scenario and specifies crop

management, price and soil data input files, units of wealth, and options for duration of the












scenario and the number of times it is replicated. The next two sections--OUTPUTS and

ANALYSES--control file and graphical output. These are discussed later in this chapter.



Table 4-1. Sections in the farm scenario file.
Section Purpose
SCENARIO Specifies title, units, other files, and simulation control.
OUTPUTS Controls farm-level file output.
ANALYSES Controls graphic display and file output.
RESOURCES Identifies farm resources.

LINKAGES Defines interrelationships between resources.

OPERATION Specifies types of operations, priorities, a time window, and
REQUIREMENTS timed resources required.
SCHEDULED Schedules operations which are not included in the experiment
OPERATIONS file or returned by crop models.

LANDSCAPE Identifies homogeneous fields and their positions and
characteristics.

LIVESTOCK Notyet developed. When developed, will specify herd and
management information for grazing livestock.
STRATEGIES Specifies sequences of management activities.
ENTERPRISES Links management (i.e., strategies) with landscape (i.e.,
subfields).
PRODUCTION Not yet developed. When developed, will specify a model and
DECISIONS criteria for selecting strategies for each enterprise.

CONSUMPTION Specifies household subsistence and discretionary consumption
DECISIONS of resources.




The remaining sections of the scenario file define both the structure and the initial


conditions of a model of a particular farm.










Resources. The RESOURCES section specifies the particular resources that

constitute a farming system's state variables, and determines their initial characteristics and

supply. A hierarchy of nine resource classes is available in FSS to represent the various

types of farm resources (Table 4-2, Fig. 4-3). Two additional abstract ancestral classes

are used internally as templates for the other classes.


Resource


Simple
resource



Consumable Timed Credit
resource resource resource



Seasonal Capital Activity-linked
resource resource credit resource



Machine
resource


Figure 4-3. Tree of resource class hierarchy in FSS.


Aggregate
resource



Debt
resource




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs