Advances in agronomy volume 67

Advisory Board
Martin Alexander

Ronald L. Phillips

Cornell University

University of Minnesota

Kenneth J. Frey

Larry P. wilding

Iowa State University

Texas A&M University

Prepared in cooperation with the

American Society of Agronomy Monograpbs Committee
Jon Bartels
Jerry M. Bigham
Jerry L. Hatfield
David M. Kral

Diane E. Stott, Chair
Linda S. Lee
David M. Miller
Matthew J. Mom

Donald C. Reicosky
John H. Rechcigl

Wayne I? Robarge
Dennis E. Rolston
Richard Shbles
JeffreyJ. Volenec


Edited by

Donald L. Sparks
Department of Plant and Soil Sciences
University of Delaware
Newark, Delaware

ACADEMIC PRESS
SanDiego London Boston NewYork Sydney Tokyo Toronto


This book is printed on acid-free paper. 63
Copyright 0 1999 by ACADEMIC PRESS
All Rights Reserved.
No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or any information
storage and retrieval system, without permission in writing from the Publisher.
The appearance of the code at the bottom of the first page of a chapter in this book
indicates the Publisher’s consent that copies of the chapter may be made for
personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay the stated per copy fee through the Copyright Clearance

Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923), for copying
beyond that permitted by Sections 107 or 108 of the U.S.Copyright Law. This consent
does not extend to other kinds of copying, such as copying for general distribution, for
advertising or promotional purposes, for creating new collective works, or for resale.
Copy fees for pre-1999 chapters are as shown on the title pages. If no fee code
appears on the title page, the copy fee is the same as for current chapters.
0065-2113/99 $30.00

Academic Press
a division of Harcourt Brace & Company
525 B Street, Suite 1900, San Diego, California 92101-4495, USA


Academic Press
24-28 Oval Road, London NW17DX. UK
/>International Standard Book Number: 0- 12-000767-3
PRINTED IN THE UNITED STATES OF AMERICA
99 00 01 02 03 M B B 9 8 7 6 5

4

3 2

1


Contents
CONTRIBUTORS
...........................................
PREFACE

.................................................

vii

ix

ASPECTS OF PRECISION
AGRICULTURE
Francis J. Pierce and Peter Nowak
I. Introduction.. ............................................

II. Overview of the Basic Components of Precison Farming. .........
III. Conclusions ..............................................
References ...............................................

2
5
65
67

SURFACE
CHARGE
AND SOLUTE
INTERACTIONSINSOILS
N. S. Bolan, R. Naidu, J. K. Syers, and R. W. Tillman
I. Introduction.. ............................................
II. Types of Electrical Surface Charge. ...........................
III. Development of Surface Charge. .............................
N. Components of Surface Charge ..............................
V. Solution-Surface Interface ..................................

VI. Concepts of Point of Zero Charge ............................
VII. Measurement of Surface Charge. .............................
W I . Factors Affecting Surface Charge .............................
M. Effect of Surface Charge on Soil Properties. ....................
X. Manipulation of Surface Charge to Control Solute Interactions. . . . .
XI. Conclusions and Future Research Needs .......................
References ...............................................

88
90
90
96
97
104
107
112
120
126
130
131

ALLELOPATHY:
PRINCIPLES,
PROCEDURES,
PROCESSES,
AND PROMISES
FOR BIOLOGICAL
CONTROL
Inderjit and K. Irwin Keating
I. Introduction. .............................................

11. SomeConcerns.. .........................................
III. Allelopathy in Agroecosystems ...............................

142
145
148


vi

CONTENTS

W. Factors Influencing Allelopathy ..............................
v. Secondary Metabolites with Allelopathic Potential . . . . . . . . . . . . . . .
VI. Mechanisms of Action of Allelopathic Chemicals . . . . . . . . . . . . . . . .
w. Allelopathic Growth Stimulation .............................
vm. Roles for Allelopathy in Biocontrol Programs ...................
Ix. Additional Comments ......................................
X . Concluding Remarks .......................................
References ...............................................

166
181
184
189
190
195
206
207

TURFGRASS
MOLECULAR
GENETICIMPROVEMENT
FOR ABIOTI~/EDAPHIC
STRESS RESISTANCE

R . R. Duncan and R. N. Carrow
I. Introduction ..............................................

TI. Molecular Genetic Improvement .............................
III. Enhancement Strategy for Multiple-Stress Resistance ............
W. Summary ................................................
References ...............................................

INDEX...................................................

233
235
275
282
283

307


Contributors

Numbers in p n t h e s z s indicate the pages on which the authors' conbibutions begin.

N. S. BOLAN (87), Department of Soil Science, Massq University, Palmerston

North $301, New Zealand
R. N . CARROW (233), Department of Crop and Soil Sciences, University of Georgia at Gnfin, &fin, Georgia 30223
R. R. DUNCAN (23 3), Department of Crop and Soil Sciences, Universityof Georgia at &fin, &fin, Georgia 30223
INDERJIT (141),Department ofAgriCultzlra1Sciences (weed Science), The Royal
Veterinury and Agricultral Univenity, DK-1871 Freabiksberg C., Copenhagen,
Denmark
K. IRWIN KEATING (141),Department ofEnvironmenta1Science, Rutgers, The
State University $New Jm-ey,New Brunswick, New 'jersey 08903
R. NAIDU (87), CSIRO Land and Water, Glen Omond 5064, South Australia,
Australia
PETER NOWAK (l), Department of Rural Sociology, University of Wisconsin,
Madison, Wisconsin 53706
FRANCISJ . PIERCE (I),Department of Crop and Soil Scimes, Michigan State
University, East Lansing, Michigan 48824
J. K. SYERS (87), Department of Agrinrltzlral and Environmental Science, University of New Castle upon Tpe, New Castle upon Tpe NE1 7RU, United Kingdom
R. W. TILLMAN (87), Department ofsoil Science, Massq University, Palmnston North 5301, New Zealand

vii


This Page Intentionally Left Blank


Preface

Volume 67 contains four comprehensiveand timely reviews of topics that should be
of great interest to professionals and students in crop and soil sciences. Chapter 1
addresses one of the most active areas in agronomic research-precision agriculture.
All aspects of the topic, including technologies, management, and economic and
environmental impacts, are discussed. Chapter 2 is a thoughtful review of surface

charge and solute interactions in soils. In addition to a theoretical treatment of the
topic, practical applications,including surface charge effects on solute interactions
and dispersiodflocculation and manipulation of surface charge by amendment additions, are included. Chapter 3 is a useful review of a topic of great interest to
agronomists-allelopathy. Principles, procedures, processes, and promises for
biological control are discussed. Chapter 4 thoroughly covers advances in the use
of molecular genetics to enhance abioticledaphic stress resistance in turfgrass.
Many thanks to the authors for their first-rate reviews.

DONALD
L. SPARKS

ix


This Page Intentionally Left Blank


ASPECTS OF PRECISION
AGRICULTURE
Francis J. Pierce' and Peter Nowak*
'Department of Crop and Soil Sciences
Michigan State University
East Lansing, Michigan 48824
2Department of Rural Sociology
University of Wisconsin
Madison, Wisconsin 53706

I. Introduction
A. Definition of Precision Agriculture
B. Intuitive Appeal

11. Overview of the Basic Components of Precision Fanning
A. The Enabling Technologies
B. Steps in Precision Agriculture
HI. Conclusions
References

Precision agriculture is the application of technologies and principles to manage
spatial and temporal variabilityassociated with all aspects of agricultural production for
the purpose of improvingcrop performance and environmental quality. Success in precision agriculture is related to how well it can be applied to assess, manage, and evaluate the space-time continuum in crop production. This theme is used here to assess the
current and potential capabilities of precision agriculture. Precision agriculture is technology enabled. It is through the integration of specific technologies that the potential
is created to assess and manage variability at levels of detail never before obtainable
and, when done correctly,at levels of quality never before achieved.The agronomic feasibility of precision agriculture has been intuitive, depending largely on the application
of traditional management recommendations at finer scales, although new approaches
are appearing. The agronomic success of precision agriculture has been limited and inconsistent although quite convincing in some cases, such as N management in sugar
beet (Bera vulgaris L.).Our analysis suggests prospects for current precision management increase as the degree of spatial dependence increases, but the degree of difficulty in achieving precision management increases with temporal variance. Thus, management parameters with high spatial dependence and low temporal variance (e.g.,

I
A&mctx in Agronomy, Volume 67

Copyright8 1999 by Academic Prea. All rights of reproduction in m y farm reserved.
0065-21 13/99 $30.00


2

FRANCIS J. PIERCE AND PETER NOWAK
liming, P, and K)will be more easily managed precisely than those with large temporal
variance (e.g.. mobile insects). The potential for economic, environmental, and social
benefits of precision agriculture is complex and largely unrealized because the spacetime continuum of crop production has not been adequately addressed.
0 1999 Academic Press

I. INTRODUCTION
It would be a simple matter to describe the earth’s surface if it were the same
everywhere. The environment,however, is not like that: there is almost endless
variety.
-Webster and Oliver ( I 990)
The quote by Webster and Oliver (1990) is particularly applicable because precision agriculture is concerned with the management of variability in the dimensions of both space and time. Without variability, the concept of precision agriculture has little meaning (Mulla and Schepers, 1997) and would never have
evolved. It appears that any component of production agriculture-from natural
resources to plants, production inputs, farm machinery, and farm operators-that
is variable in some way is included in the realm of precision agriculture. Aspects
of precision agriculture, therefore, encompass a broad array of topics, including
variability of the soil resource base, weather, plant genetics, crop diversity, machinery performance, and most physical, chemical, and biological inputs used in
the production of a crop, whether natural or synthetic. By necessity, these aspects
are all framed within the context of the socioeconomicaspects of production agriculture because to be successful on the farm, precision agriculture must fit the
needs and capabilities of the farmer (Nowak, 1997) and must be profitable (Lowenberg-DeBoer and Swinton, 1997).
Bell et al. (1995) state correctly that efforts toward precision agricultural management should recognize that the factors affecting crop yields and environmental sensitivity vary in both space and time. Managing soils and crops in space and
time is the sustainable management principle for the twenty-first century, a principle exemplified by farming by soilscapes, managing zones within the field, and
managing the noncrop period (Pierce and Lal, 1991). The unifying theme of this
chapter is that success in precision agriculture is directly related to how well it can
be applied to manage the space-time continuum in crop production. We postulate
that prospects for precision management increase as the degree of spatial dependence increases, but the degree of difficulty in achieving precision management
increases with temporal variance. Thus, for management parameters that vary spatially, those with high temporal correlations (e.g., liming) will be more easily man-


ASPECTS OF PRECISION AGRICULTURE

3

aged with precision agriculture than those with large temporal variance (e.g., mobile insects). Within a given management parameter, the success to date of precision management is to a large extent determined by the degree to which the spatial variability is temporally stable.
This chapter provides an overview of precision agriculture and an assessment

of its current state and its potential to improve crop performance and environmental quality in production agriculture. In this chapter, we define precision agriculture, explore the technological capabilities that enable it, assess its agronomic
feasibility and environmental efficacy, and evaluate its performance to date relative to economic and social impacts. The chapter concludes with an analysis to
identify needed developments in precision agriculture and we provide some
thoughts for a future research agenda. Given the expansive nature of precision
agriculture coupled with space constraints, we attempt to synthesize the important
aspects of precision agriculture while guiding the reader to the growing volume of
literature on the subject. Readers seeking more detail are referred to the following
major publications related to precision agriculture: Auernhammer ( 1994),American Society of Agricultural Engineers (ASAE) (1991), BIOS (1997), Lake et al.
(1997), National Research Council (NRC) (1997), Pierce and Sadler (1997),
Robert et al. (1993, 1995, 1996), Sawyer (1994, Schueller (1992), Stafford
(1996b), and Sudduth (1998). We are aware of the rapid rate of change in precision agriculture and the inadequacies this causes in an overview of this nature.

A. DEFINITION
OF PRECISION
AGRICULTURE
Currently, no precision agricultural systems exist; rather, various components
of traditional crop management systems have been addressed separately regarding
their potential for site-specific management, perhaps most notably soil fertility
(Lowenberg-DeBoer and Swinton, 1997).The state of precision agriculture from
a systems perspective is analogous to the early days of no-tillage crop production.
Technology became available in the 1960sto plant seeds in untilled soil, but it was
not until the many aspects of crop production were adequately addressed under
lack of tillage and crop residue management, including the management of fertility and pests, that successful no-tillage systems were developed and implemented
(Blevins et al., 1998). The adoption of no-tillage did not proceed at a significant
rate until the 1980s when the integration of appropriate technologies and public
policies supported its dissemination to farmers (Allmaras et al., 1998; Larson et
al., 1998;Now& and Korsching, 1998).In a similar fashion, although certain technologies in the early days of precision agriculture allowed for the variable application of nutrients and pesticides, there did not exist a thorough understanding of
how soil fertility and pests varied in space and time. Most important, explanations
were lacking on what specifically caused the variability so that appropriate inputs

4

FRANCIS J. PIERCE AND PETER NOWAK

could be matched to site-specific conditions. Today, farmers are adopting individual components of precision agriculture on the farm but a distinctive precision
farming system has not yet evolved. Technological developments continue to occur and as a result of ongoing research a better understanding of underlying
processes is being developed but a true system has not emerged. Therefore, any
definition of precision agriculture can at best be considered only operational.
Since the mid-l980s, a host of terms have been used to describe the concept of
precision agriculture, including farming by the foot (Reichenberger and Russnogle, 1989), farming by soil (Carr et al., 1991; Larson and Robert, 1991). variable rate technology (VRT) (Sawyer, 1994), spatially variable, precision, prescription, or site-specific crop production (Schueller, 1991), and site-specific
management (Pierce and Sadler, 1997). All these terms, however, have in common the concept of managing variability at scales that are within fields. As
Stafford (1996b; p. 595) states, precision agriculture involves “the targeting of inputs to arable crop production according to crop requirements on a localized basis.” Thus, the intent of precision agriculture is to match agricultural inputs and
practices to localized conditions within a field to do the right thing, in the right
place, at the right time, and in the right way (Pierce ef al., 1994). A recent report
of a National Research Council, Board on Agriculture Committee defined precision agriculture as “a management strategy that uses information technologies to
bring data from multiple sources to bear on decisions associated with crop production” (NRC, 1997; p. 17). While the NRC definition raises important informational dimensions of precision agriculture, it fails to emphasize the basic
premise of precision agriculture-the management of spatial and temporal variability. In this chapter, we use the following definition of precision agriculture as
the basis of our discussions: Precision agriculture is the application of technologies and principles to manage spatial and temporal variability associated with all
aspects of agricultural production for the purpose of improving crop performance
and environmental quality.
We provide a final note on the word precision because there is sure to be confusion regarding its meaning in precision agriculture versus its use in statistics. The
term precision refers to the quality or state of being precise, where precise means
minutely exact, a term synonymous with correct. Precision agriculturerefers to exactness and implies correctness or accuracy in any aspect of production. In statistics, however, precision is the closeness of repeated measurements of the same
quantity to each other, whereas accuracy is the closeness of a measured or computed value to its true value (Sokal and Rohlf, 1995). In measurements, accuracy
is synonymous with correctness (i.e., validity), whereas precision refers to reproducibility (i.e., reliability). Thus, something can be precise but not accurate. Another matter is measurement precision implied by number of digits reported for a
given measurement. The nature of computers makes it easy to imply more precision than was possible in various aspects of data collection, analysis, and compu-


ASPECTS OF PRECISION AGRICULTURE

5

tation in precision agriculture. Precision here refers to the limits on the measurement scale between which the true measurement is believed to lie, implied by the
number of digits reported for a measurement (Sokal and Rohlf, 1995). The more
digits reported for a measurement, the higher the precision implied. A pH of 5.44
implies more precision than a pH of 5.4. The appropriate precision with which to
report a number is to include one additional digit beyond the last significant one
measured by the observer. Statistics plays an important role in the application of
precision agriculture and care should be taken in dealing with accuracy, precision,
and implied precision in the reporting data.

B. INTUITIVEAPPEAL
Precision agriculture is intuitively appealing because it is closely aligned with
the scientific principles of management of soils, crops, and pests. Few would argue against a management philosophy that espouses matching inputs to the exact
needs everywhere. Precision agriculture is intuitively appealing because it offers
a means to improve crop performance and environmental quality in production
agriculture (Wolf and Nowak, 1995). While the intuitive appeal creates high expectations for precision agriculture, the physical evidence supporting the agronomic (Lowenberg-DeBoer and Swinton, 1997; Sawyer, 1994) and environmental (Larson et al., 1997) benefits of precision agriculture is limited in part because
it is still in its infancy.
As we will demonstrate in our discussion, successful implementation of precision agriculture depends on numerous factors, including (i) the extent to which
conditions within a field are known and manageable, (ii) the adequacy of input recommendations, (iii) the degree of application control, and (iv) the degree of support through private and public infrastructures. Individual success also depends on
the expectationsplaced on precision agriculture which represent the difference between promotional and educational efforts versus the actual experience of farmers.

II. OVERVIEW OF THE BASIC COMPONENTS
OF PRECISION FARMING
The main componentsof any precision agriculture system that may emerge must
first address the measurement and understanding of variability. Next, this system
must use information to manage this variability by matching inputs to conditions
within fields using site-specific management recommendations and mechanisms
to control the accuracy of site-specific inputs. Finally, and most important, this system must provide for the measurement and recording of the efficiency and effica-

6

FRANCIS J. PIERCE AND PETER NOWAK

cy of these site-specific practices in order to assess value on and off the farm. Thus,
precision agriculture is technology enabled, information based, and decision focused (Pierce, 1997a).

A. THEENABLING
TECHNOLOGIES
While the concept of matching inputs to site-specific conditions is not new, as
just discussed, there is little doubt that important advances in technology continue to enable precision agriculture. The enabling technologies of precision agriculture can be grouped into five major categories: computers, global position system
(GPS), geographic information systems (GIS), sensors, and application control.
Few of the enabling technologies were developed specifically for agriculture and
their origins date back more than 20 years, as illustrated in the time chart in Figure 1. It is the integration of these technologies that has enabled farmers and their
service providers to do things not previously possible, at levels of detail never
before obtainable, and, when done correctly, at levels of quality never before
achieved (Fortin and Pierce, 1998).
1. Computers

Many technologies support precision agriculture, but none is more important
than computers in making precision agriculture possible. Also, it is not computers
alone that are important but their ability to communicate that is so powerful for
agriculture.As Taylor and Wacker (1997) suggest, it is the fusion of computers and
communication that gave birth to connectivity, and it is connectivity that is driving
the access of everyone to everyone, everything to everything, and everything to
everyone. This electronic linkage and communication define the age of access
(Taylor and Wacker, 1997). It is this notion that may have prompted the NRC
(1997) to define precision agriculture in terms of a management strategy that uses

information technologies for decision making.
Precision agriculture requires the acquisition, management, analysis, and output of large amounts of spatial and temporal data. Mobile computing systems were
needed to function on the go in farming operations because desktop systems in the
farm office were not sufficient. These mobile systems needed microprocessorsthat
could operate at speeds of millions of instructions per second (MIPS), had expansive memory, and could store massive amounts of data. The first microchip created by Intel in 1971 (Intel 4044 processor) contained a mere 2300 transistors and
performed about 60,000instructions per second. Since 1971, the number of transistors per chip has doubled every 18 months (Fig. 2) affirming Gordon Moore’s
observation in 1965 that a doubling of transistor density on a manufactured die
was occurring every year, a concept referred to as “Moore’s law” (Moore, 1997).


Date

Event

1840s

Aerial photography emerges; pictures taken from balloons
First image sensors incorporated in satellites; low-resolution black and white TV
First commercial CIS
First chlorophyll sensor (Benedict and Swidler, 1961)
First multispectral photography done from space Apollo 9 manned mission
Baumgardener et al. (1970) related soil organic matter to multispectral data
Intel 4040 processor
Launch of Earth Resources Technology Satellite-1 later renamed Landsat; permitted
continuous coverage of most of the earth's surface
Soil organic matter sensor (Page, 1974)
Apple computer commercialized ()

1960s
1961

1968
1970
1971
1972
1974
1977
1978
1980
1981
1982

1983
1984

1985
1986

1987

1988

I990
1991
1992
1993
I994
1996
1997

Launch of first NAVSTAR GPS satellite

First IBM PC
Intel 80286 processor
Launch of Landsat "Thematic
Mapper ("h4) added
The Jet Propulsion Lab produces hyperspectral sensors for use from a high-altitude
aircraft platforms known as AIS (Airborne Imaging Spectrometer)
GPS available for civilian use
Ortlip patent issued to SoilTEQ
Launch of Landsat 5
286 Intel processor
Grain flow monitoring on combines @e Baerdwmeker et al., 1985)
French launch an operational series of earth-observingsatellites called SPOT (SysPme
Probatoire d'observation de la Terre); first offering of multispectral data to world
users on a commercial basis
The Jet Propulsion Lab produces a second hyperspectral sensor known as AVIRIS
(Airborne VisibldnfraRed Imaging Spectrometer)
Yield mapping in Texas (Bae et al.. 1987)
India launches earth resources satellite (IRS-IA) that gathers data in the visible and
near IR with the Linear Imaging Self-scanning sensor (LISS)
Intel 40486 processor
Canadian Radarsat, ERS-I, and ERS-2 managed by the European Space Agency A
class of satellite remote sensors using radar systems
Japan launches JERS- 1 and JERS-2 that include both optical and radar sensors
Selective availability (SA) imposed on GPS signal
First symposium on site-specific crop production (ASAE, 1991)
Commercial yield monitors appear in the United States
First international conference on soil specific crop management (Robert et al., 1993)
Pentium processor
Full constellation of 24 GPS satellites in NAVSTAR system complete
Earth System Science Pathfinder launched by NASA

Pentium I1 processor
India launches the latest in the series. IRS-ID, on September 29, 1997
First European conference on precision agriculture (BIOS, 1997)
Board of Agriculture, National Research Council report on precision agriculture
(NRC,1997)

Figure 1 Historical developments in the technologies that enabled precision agriculture.


8

FRANCIS J. PIERCE AND PETER NOWAK

Yorr

Figure 2 Illustration of Moore’s law showing the doubling of computer speed and capacity every
year [Source:Intel Corporation (www.intel.com)].

As Fig. 2 indicates, Moore’s law is expected to hold until 2017 (according to
Moore) and appears to hold for memory and storage capacity. Data storage capacity will need to increase rapidly as sensor technology and digital geospatial data
become increasingly available to agriculture. Moore notes,
By the Year 2012, Intel should have the ability to integrate 1 billion transistors
onto a production die that will be operating at 10 GHz. This could result in a
performance of 100,OOOMIPS, the same increase over the currently cutting edge
Pentium I1 processor as the Pentium 11processor was to the 386! We see no fundamental barriers in our path to Micro 2012, and it’s not until the Year 2017 that
we see the physical limitations of wafer fabrication technology being reached.
We can expect, therefore, that computers will drive significant technological development to enable precision agriculture for the foreseeable future. The extent to
which agriculture can utilize computer technology is important to the success of
agriculture in general (Holt, 1985; Ortmann et al., 1994). However, the agricultural sector is lagging in the adoption of computer technologies on the farm relative to other business sectors. According to the 1997 annual survey of the U.S. Department of Agriculture (USDA) National Agricultural Statistics Service (NASS,
1997), of 2,053,800 farms in the United States, only 38% had computer access,

31% owned or leased a computer, and 13% had Internet access. Part of this computer lag in agriculture is due to the lack of access or connectivity in rural areas,
lack of training, and little perceived utility in available software (Nowak, 1997;


ASPECTS OF PRECISION AGRICULTURE

9

Peterson and Beck, 1997).In any event, farmers will have to become as comfortable working with computers and their data as they are working with their farm
machinery (Klein, personal communication, 1997).
While it appears that computer hardware will be more than adequate for precision agriculture, the same cannot be said for the software. Advances in software
logically lag behind the hardware technology. However, software for precision
agriculture has been more an experience than an application. Berry (1999, in discussing the human factor in GIS, describes experience as “what you get when you
don’t get what you want.” Computer software in precision agriculture has become
better with time, but precision agriculture is loaded with Berry’s type of experience. Software will be adequate for precision agriculture when it becomes, as
Berry (1 995) suggests, second nature to the user for assessing information and
translating it into knowledge. For precision agriculture, the knowledge needed is
that for managing variability on the farm, knowledge that is requisite for decision
making. Computers and salient, usable software are going to play a critical role in
the emergence of a precision agriculture system in the near future.

2. Geographic Information Systems
Formally, GIS is an organized collection of computer hardware, software, geographic data, and personnel designed to efficiently capture, store, update, manipulate, analyze, and display all forms of geographicallyreferenced information [Environmental Systems Research Institute (ESRI), 19971. The GIS concept dates
back to the 1960s when computersbecame available for use in spatial analysis and
quantitative thematic mapping (Burrough, 1986).The science of GIS has evolved
since the 1960s to include data management and modeling, enabling a shift from
mapping to spatial reasoning (Berry, 1993, 1995). The ability to perform spatial
operations on the data distinguishes a true GIS from the many software programs
that do thematic mapping and database management. During the past few years,
mapping software programs have been adding spatial operations, workstation GIS

software programs have spawned microcomputer versions with more limited GIS
capabilities to fit desktop computer technologies, and new microcomputer-based
GIS systems have emerged. There are many different mapping and GIS software
programs that offer different GIS features. None, however, have captured the market for application in precision agriculture.
Because precision agriculture is concerned with spatial and temporal variability and because it is information based and decision focused (Pierce, 1997a), it is
the spatial analysis capabilitiesof GIS that enable precision agriculture. This statement is true because the value of precision agriculture is derived only when resulting information is turned into a management decision that increases profitability, benefits the environment, or provides some other value to the farm. AGIS,
in the full sense of its formal definition given previously, is key to extracting val-


10

FRANCIS J. PIERCE AND PETER NOWAK

ue from information on variability. Clark and McGucken (1996) refer to GIS as
the brain of a precision farming system. However, available GIS software packages are complex and difficult to learn for nonspecialists. Some GIS lack data management and spatial analysis tools needed to understand the variability observed
on the farm and needed to derive site-specific management recommendations.
More functional, easy to use interfaces are needed in order to fully utilize this technology in production agriculture (Berry, 1995; NRC, 1997).Computer simulation
modeling can help derive the needed understanding of variability (Sadler and Russell, 1997; Verhagen and Bouma, 1997) and linking GIS to models (Goodchild et
al., 1993) will be important to precision agriculture.

3. Global Positioning Systems
Location control (Schueller, 1992) is essential to precision agriculture for assessing spatial variability and for site-specific application control (Auernhammer,
1994; Tyler et al., 1997). In the early days of precision agriculture, relative position within a field was determined by dead reckoning. This was a simple method
in which position was measured relative to a known point determined by measuring distance using radar, ultrasound, and wheel shaft counters. Direction was determined either by using a steering-angle sensor or a gyroscope from the known
point or by direction only if a field had linearized tramlines of known fixed location (Auernhammer and Muhr, 1991). Triangulation methods, in which position is
determined relative to two or more known locations using, for example, radio signals transmitted from reference stations to mobile units (Palmer, 1991, 1995; Scorer, 1991), improved position accuracy to as low as 15 cm (95%probability) but
such systems were time-consuming and expensive. By the early 1990s. however,
the GPS known as NAVSTAR (NAVigationSystem with Time And Ranging) was
becoming available for general civilian use including agriculture.This system was
based on 18 satellites that were in orbit by early 1990 (Hoffmann-Wellenhofet al.,

1994; Kaplan, 1996; Kennedy, 1996; Leick, 1995; NAPA, 1995). The United
States NAVSTAR GPS system consists of a constellation of 24 satellites, including 3 spares. The first satellite was launched in 1978 but it was not until the Soviet downing of a Korean airliner in 1983 that the decision was made to make GPS
available for civilian use [National Academy of Public Administration (NAPA),
19951. The NAVSTAR GPS system was fully deployed by 1994 and declared
fully operational in 1995. The Russians also deployed a GPS system called
GLONASS (Global Navigation Satellite System) consisting of 24 satellites completed in 1995. Although there are differences in time standards and coordinate
systems between GLONASS and NAVSTAR, higher end GPS receivers currently
available accommodate the combined use of both GPS systems resulting in increased reliability and accuracy. Although the Russian GLONASS policy called
for ensured availability for 15 years, no charge on a constant global basis, and no
selective availability, the system was degraded to only 14 or 15 active spacecraft


ASPECTS OF PRECISION AGRICULTURE

11

during the fall of 1997 (Perry, 1998). Therefore, changes in GPS technology are
to be expected.
The GPS technology enables precision agriculture because all phases of precision agriculture require positioning information. GPS is able to provide the positioning in a practical and efficient manner for a few thousand dollars ( v l e r et al.,
1997). Expensive, high-precision differential GPS (DGPS) systems are available
that achieve centimeter accuracies (Lange, 1996), allow for automated machinery
guidance (O’Conner et al., 1996; 51er et al., 1997) and kinematic mapping of
topography (Clark, 1996), and are useful in the creation of digital elevation models needed for terrain analysis (Bell et al., 1995; Moore et al., 1993). While the
GPS signal is ubiquitous, there have been problems in making available GPS at
the needed precision for agriculture (Saunders et al., 1996). The U.S.Department
of Defense implemented selective availability (SA) on March 25, 1990, which limited accuracy of GPS to civilians from about 8-10 m without SA to about 100 m
with SA. This was done by varying the reported precise time of clocks on board
the satellites and by providing incorrect orbital positioning data (NAPA, 1995).
The SA has been overcome by the use of differential corrections transmitted to
GPS receivers (rovers) from GPS receivers at known fixed locations (base). DGPS

involves the transmission of a differential correction, that is, the difference between actual and predicted position at the base GPS receiver, to rover GPS receivers, which then apply the corrections to received GPS signals to solve for a
more accurate position (Qler et al., 1997). There are four general ways of providing a differential correction: a private local GPS base receiver with a radio modem to transmit to a mobile receiver, a commercial GPS base station at which differential corrections are transmitted on FM subcarrier frequencies, a public GPS
base station at which differential corrections are transmitted on AM frequencies
from radio beacons with up to a 250-mile radius [U.S. Coast Guard (USCG) beacon system), and a wide area differential GPS (WADGPS) network in which differential corrections from a network of base stations are used by the roving GPS
receiver to correct its position (vier et al., 1997). In all cases, DGPS requires additional receivers and antennas and is fee based for commercial correction
providers. A differentialcorrection is desirable even without SA because it is needed to achieve the accuracies needed in some aspects of precision agriculture, including navigation and guidance. Currently, only WADGPS provides national coverage, whereas all others are dependent on whether the rover is close enough to a
base station to receive the signal consistently. However, this is changing because
FM providers are planning to offer national coverage in the near future and there
are plans for completion of the USCG beacon system nationally (Divis, 1998).
There is currently a debate as to whether the public sector should provide a national DGPS (NDGPS) to agriculture (NAPA, 1995;Pointon, 1997). Other sectors
of the U.S. economy also need a national NDGPS, so the discussion of who benefits from a publicly supported NDGPS should not be focused on agriculture alone.
The Office of Management and Budget did not support expansion of the USCG


12

FRANCIS J. PIERCE AND PETER NOWAK

radio beacon system for public NDGPS in Ey98. However, some believe that a
government-provided NDGPS system is so important to critical activities that it is
best for the government to provide it (Divis, 1998). Certainly, precision agriculture needs DGPS and will require increased position accuracy as new technologies for navigation and guidance require higher precision, which may require
DGPS accuracies not available from a government NDGPS. The prophecy of
Auernhammer and Muhr (1991; p. 395) that “their use will also be without costs
in the future” will probably never be realized because DGPS is big business.
Regardless of who provides all aspects of DGPS, farmers and their service
providers need reliable DGPS to achieve the desired positioning for precision
farming operations. Farmers still experience interruptions and interferences in the
GPS and/or differential correction signals, creating gaps in data collection or loss
of application or guidance control. In activities at higher speeds, such as aerial applications (Kirk and Tom, 1996), time delays in differential corrections may limit
positional accuracy in kinematic mode (NRC, 1997). The specified availability

(four satellites in view at any location) of the NAVSTAR system is 99.85%, with
a reliability (system is in service when it needs to be) of 99.97% (NAPA, 1995).
However, the suitability of the satellite geometry for calculating a solution, referred to as dilution of precision (DOP), is a problem in farming in which natural
or man-made structures obstruct the receivers’ view of some satellites or interfere
with differential correction reception. There are also geographic locations at which
DOP has been inadequate for needed location precision at certain times during the
day. Additionally,some GPS receivers are susceptible to unwanted interfering signals from a variety of sources, including farm machinery, making the receiver useless in navigation or positioning. Some interferences can be overcome in the design of the GPS receiver.
Regardless of problems, DGPS has greatly enabled precision agriculture. Of
great importance for precision agriculture, particularly for guidance and for digital elevation modeling, position accuracies at the centimeter level are possible in
DGPS receivers that use carrier phase in combination with DGPS (Lange, 1996;
Tyler er al., 1997).Accurate guidance and navigation systems will allow for farming operations not currently in use, including field operations at night when wind
speeds are low and more suitable for spraying and the use of night tillage to reduce the light-induced germination of certain weeds (Hartmann and Nezadal,
1990). DGPS technology changes continually and can be followed on the internet
(e.g., Peter Dana’s web site hrtp://wwwhost.cc.utexus.edu/Stp/pub/grg/gcrafr/
notes/gps/gps.html or www.gpsworld.com).

4. Sensors
Sensors are devices that transmit an impulse in response to a physical stimulus
such as heat, light, magnetism, motion, pressure, and sound. With computers to


ASPECTS OF PRECISION AGRICULTURE

13

record the sensor impulse, a GPS to measure position, and a GIS to map and analyze the sensor data, any sensor output can be mapped at very fine scales. Sensor
technology currently lags behind other enabling technologies (Sudduth et al.,
1997) and the availability of sensors has been cited as the most critical factor preventing the wider implementation of precision agriculture (Stafford, 1996b). Sensors are critical to success in the development of a precision agricultural system
for three important reasons: Sensors have fixed costs, sensors can sample at very
small scales of space and time, and sensors facilitate repeated measures. This

means that the cost per sample is determined by the extent of sensor use, sample
intensity is determined by the capability of the sensor and not the cost or difficulty in sampling associated with traditional physical sampling schemes, and sampling frequency is determined by accessibility of the target and not costs.
The value of sensors and their potential for the future of precision agriculture are
illustrated by yield monitoring.Yield monitoring systems, which use sensorsto measure crop flow, allow the creation of yield maps with detail not practical with other
measurementtechniques (Pierce er al., 1997).Yield mapping technology may be the
major factor responsible for the growing interest in precision agriculture observed
since its commercial introduction in 1992 (Stafford, 1996b). Prior to 1992, the focus was on VRT, which would not in itself have sustained precision agriculture.
Yield mapping bolstered precision agriculture and is currently the major precision
agriculture technology in U.S. agriculture. However, the promise of sensing technologies may make yield mapping technology unnecessary in the future if high-resolution remote sensing of the growing crop leads to quantitative prediction of crop
yield prior to harvest. Yield mapping will serve to validate sensor-based predictive
technology, but once operational, yield monitors may not be needed. The use of remote sensing to forecastcrop yields is in use worldwide, and forecastingoffers farmers the ability to market their crops prior to harvest when prices are more favorable.
Sensors are very desirable for use in precision agriculture. Every effort should
be made to promote the application and adaptation of sensors developed in other
industrial sectors, especially the space and defense industries, as well as to promote the development of new sensor technologies for use in assessing and managing variability in soils, plants, pests, and machinery. Sensors can be contact or
remote, ground based or space based, and direct or indirect. Sensors have been developed to measure machinery, soil, plants, pests, atmosphericproperties, and water by sensing motion, sound, pressure, strain, heat, light, and magnetism and relating these to properties such as reflectance, resistance, absorbance, capacitance,
and conductance. Sensors are needed in precision agriculture because such a system requires the collection,coordination, and analysis of massive quantities of data
(Sudduth et al., 1997),some for strategic surveys and inventories and some for use
in real-time applications.
Remote sensing involves the detection and measurement of photons of differing energies emanating from distant materials. These photons may be identified


14

FRANCIS J. PIERCE AND PETER NOWAK

and categorized by classhype, substance, and spatial distribution, with most designed to monitor reflected radiation (Frazier et al., 1997). Satellite remote sensing dates back to the first aerial photographs taken from balloons in the 1840s.The
first satellite imagery was obtained from TV cameras mounted in satellites in the
early 1960s. Since the U.S.Landsat program launched the first observation satellite in 1972, earth observation has increased and currently India, France, Russia,
Japan, and the European Space Agency also operate earth observation satellites
(Figure 1). Many companies now offer commercial products to agriculture from

images obtained by these satellites or enhanced digital products derived from
them. Remote-sensing satellites collect image data actively by sending a known
signal from the satellite to the earth and measuring the portion of the signal that is
returned. Passive data collection occurs by measuring the incoming energy from
the sun reflected by an object or heat energy emanated from an object. The electromagnetic energy emanated from an object varies in wavelengths as determined
by the object’s physical and chemical structure. Different images of an object can
be constructed by combining different wavelengths, creating images far more revealing than images obtained from visible light alone. Remote-sensing systems
vary in spatial resolution (meters to kilometers), spectral coverage (portion of the
light spectrum covered), and temporal frequency (days to months). Different applications in agriculture will require different spatial resolutions, spectral coverages, or temporal frequencies. NASA (1998) provides an online tutorial on remote
sensing and its applications. Moran et al. (1997) provide a comprehensive review
of image based remote sensing for precision agriculture.
Remote sensing holds great promise for precision agriculture because of its potential for monitoring spatial variability over time at high resolution (Hatfield and
Pinter, 1993; Moran e? al., 1997; Stevens, 1993). For example, monitoring of a
growing crop using remote sensing is critical because yield maps document yield
variability but do not provide information on the cause of observed variability.
However, the promise of remote sensing for agriculture has not been realized for
many reasons, including costs, timeliness, and availability (Frazier et al., 1997;
Stafford, 1996b).

5. Application Control
Control is that portion of an automated system in which sensed information is
used to influence the system’s state in order to meet an objective (Stone, 1991).
For precision agriculture, control must be achieved in space and time for varying
single or multiple inputs at different rates, at varying soil depths, and in a uniform
and location-specific manner within fields. Because it is a required component,
control technology has been a strength of precision agriculture since its inception
and the state of application control was recently reviewed by Anderson and Humburg (1 997). Simply stated, if the needed accuracy cannot be achieved at the point