National Academies Press: OpenBook
« Previous: 3 Adoption of Precision Agriculture
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 90
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 91
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 92
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 93
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 94
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 95
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 96
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 97
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 98
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 99
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 100
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 101
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 102
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 103
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 104
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 105
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 106
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 107
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 108
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 109
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 110
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 111
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 112
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 113
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 114
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 115
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 116
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 117
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 118
Suggested Citation:"4 Public Policy and Precision Agriculture." National Research Council. 1997. Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management. Washington, DC: The National Academies Press. doi: 10.17226/5491.
×
Page 119

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4 Public Policy and Precision Agriculture PURPOSES FOR PUBLIC INVOLVEMENT The development and diffusion of precision agriculture has proceeded with little explicit public input. Precision agriculture developed by assembling the com- ponents of technologies developed for purposes far removed from agricultural production. The private sector made a significant investment to tailor information technologies for agricultural applications, but the public sector agricultural re- search community contributed relatively little. It is anticipated that private sector investments in development and diffusion of precision agriculture will continue at a rapid pace. This chapter discusses roles appropriate to the public sector. The array of possible public sector roles ranges from research, education, and exten- sion, to development of infrastructure for communications and institutional un- derpinnings for intellectual property rights. The committee examines the pros and cons of public participation in these roles, focusing on roles that provide public goods not likely to be addressed by private sector initiatives. Despite interest by several pioneering Agricultural Research Service (ARS) scientists concerned with agricultural engineering uses for the global positioning system (GPS) and geographic information systems (GIS), precision agriculture has not been the subject of major U.S. Department of Agriculture (USDA) re- search initiatives. Since 1990, both research and publicly funded extension efforts have been criticized for lagging behind the explosive development of precision agriculture. Innovative producers who were early adopters complain that they teach extension and research personnel and cannot obtain the research results and comparative field data that would validate claims of competing precision agriculture purvey- ors or systems. For example, little research has been conducted on interactions 90

PUBLIC POLICY AND PRECISION AGRICULTURE 91 between soil characteristics that affect fertility recommendations—such as rela- tive field elevation, nitrogen content, carbon content, and soil moisture—despite the recognition of substantial subfield variation in expected yield (Huggins and Alderfer, 1995; Larson et al., 1997; Pan et al., 1997; Vetsch et al., 1995). Thus far, the public sector has primarily contributed to precision agriculture indirectly through large infrastructure investments outside of agriculture. The largest and most critical investment has been in GPS development and implemen- tation, which was motivated by U.S. Department of Defense (DOD) needs for accurate and instantaneous navigational positioning across the world. The $12 billion invested since 1990 largely overlooked potential civilian spin-off applica- tions, including those in agriculture (National Research Council, 1995c). Other public investments in remote sensing systems, particularly the National Aeronau- tics and Space Administration LANDSAT sensors, were at least partly motivated by potential applications in agriculture. Public investments in defense computer networks such as ARPANET, leading up to development of the Internet, ben- efited the entire computing and communications community, including agricul- ture. Other federal agencies such as the U.S. Department of Energy (DOE), the U.S. Environmental Protection Agency (EPA), and the U.S. Geological Survey also provided technical expertise to USDA relevant to precision agriculture. Pri- vate industry has made large investments in these technologies, as well, often leveraged on these public investments. In 1995 the USDA Agricultural Research Service had $4.4 million directly invested in precision agriculture research projects at 15 locations (Agricultural Research Service, 1995). A general survey of ARS researchers done in mid-1996 showed 125 full-time-equivalent staff and $26 million in research activities gen- erally related to precision agriculture, about half of which was directly related to precision agriculture topics and half to supportive research. Another 45 full-time- equivalent staff and $9 million were reported as Cooperative State Research, Edu- cation, and Extension Service (CSREES) funding for precision agriculture re- search to the land grant universities. However, many of the activities reported are only partially associated with precision agriculture and cannot be accurately sepa- rated from other research areas, such as integrated pest management, sustainable agriculture, conventional yield research plots and experiments, water quality re- search, and soil nutrient and productivity research. As adoption of precision agriculture increases, explicit public policies could be formulated to foster or retard adoption. These should be focused on public benefits from adoption that do not compete with private industry objectives and cannot be realized exclusively by any one individual or company. An important reason for public involvement is to avoid any unintended consequences and dan- gers that might be caused by the increasingly widespread conversion to precision agriculture technologies. The committee identified ways that public involvement could be justified to further appropriate development and dissemination of preci- sion agriculture already undertaken in the private sector:

92 PRECISION AGRICULTURE IN THE 21ST CENTURY • providing information on advantages and disadvantages of precision agri- culture to potential adopters through research, development, and technol- ogy transfer and dissemination activities; • providing education and human capital infrastructure; • stimulating public data collection efforts; • providing physical infrastructure in cases where there are substantial economies of scale; and • protecting private property rights, particularly with respect to intellectual property such as software and data ownership. The committee also perceived a role for government in helping set standards for data storage and transfer because of the existence of external networks. This chapter reviews each of these rationales for government intervention and dis- cusses their applicability to precision agriculture. RESEARCH AND DEVELOPMENT The public sector in the United States has played a major role in research, development, and dissemination of new agricultural technologies for more than a century. Until the onset of World War II, agricultural research at USDA was the principal component of federal research and development efforts, accounting for almost 40 percent of all federal R&D spending (Mowery and Rosenberg, 1989). In 1991, by contrast, USDA research spending accounted for only 2 percent of federal research and development spending. Public sector agricultural research and development has changed little in real terms since 1980 (Fuglie et al., 1996). Agricultural research at the state level has been carried out by state agricultural experiment stations (SAESs) under the direction of the land grant university sys- tem created by the Morrill Act in 1862. The Extension Service was established as a cooperative venture between the state land grant colleges and USDA (National Research Council, 1995b). Today the public sector apparatus for research and development of new agri- cultural technologies consists of USDA’s research arms (Agricultural Research Service, Economic Research Service, and National Agricultural Statistics Ser- vice) and the network of SAESs. USDA helps set SAES research and develop- ment priorities through competitive grant funding overseen by CSREES. In 1992 total public research and development spending on agriculture was about $2.9 billion, or 46 percent of public and private agricultural research and development spending in the United States. Until 1978, real expenditures on agricultural research and development from private sources equaled public expenditures (Fuglie et al., 1996). Since then, pub- lic funding has remained nearly flat at about $2.5 to 3 billion, while private in- vestments have increased to nearly $4 billion. The totals mask a significant shift in emphasis in the type of agricultural research conducted by the private sector. In

PUBLIC POLICY AND PRECISION AGRICULTURE 93 1960, more than 80 percent of private research funding was for improving farm machinery or developing new food products or processing methods, while public research focused on increasing crop and livestock yields. By 1992, 60 percent of private research was also devoted to increasing crop and livestock yields by im- proving crop varieties, agricultural chemicals, animal breeding, feeds, and phar- maceuticals. These trends point toward more potential for competition between private and public agricultural research and development and a less clear-cut di- vision of labor. Most of the research and development embodied in current preci- sion agriculture technologies has come about either through public investments in defense and space technologies or by the private sector; there has been little investment in precision agriculture by traditional public agricultural research in- stitutions. There is every reason to believe that private research and development investment in precision agriculture will continue to be made as long as there is potential for profit. What is not clear are appropriate roles for public research and development in precision agriculture that are not duplicative of private efforts and that can materially improve development and adoption. Left to itself, the private sector will generally underinvest in socially desir- able research for several reasons: • Gains from research investments may be difficult to protect from com- petitors. • Basic research may be too risky to justify investment. • Potential markets for products of research may be too small. • Traditional technologies may have fully captured the market. • Available labor is not trained to use the new technologies. • Clients may have no incentives to adopt products of research, particularly those that improve environmental quality. Research and development are costly, and it is difficult for firms to appropri- ate the fruits of their research and development efforts because, once known, the results of those efforts can be copied easily and inexpensively. For example, com- petitors of a firm that has invented a new piece of equipment can reverse-engineer their own versions and thus produce equipment of equal or better capability with- out having to invest in the initial research and development. Patents and other forms of intellectual property rights were designed to en- courage the private sector to conduct research (Fuglie et al., 1996). Patent law is geared toward protecting inventions that embody new knowledge, not toward protecting what Huffman and Evenson (1993) term pretechnology science (i.e., scientific research applied to specific problem areas but not toward the develop- ment of products, inventions, or other patentable items), which is thus generally neglected by the private sector. For example, research on farming methods en- hances knowledge about crop productivity under alternative management sys- tems. Such pretechnology science cannot be patented, and the private sector has an incentive to engage in research of this kind only if the resulting knowledge is

94 PRECISION AGRICULTURE IN THE 21ST CENTURY expected to increase sales of a particular product, such as a specific agricultural chemical or piece of farm equipment. At the same time, knowledge of this kind is too specific to be of interest from a general scientific point of view and thus tends to be neglected by purely academic researchers as well. Where the private sector does not have incentives to conduct such research, public sector research and development may be required to fill the gaps. In some cases, research and development may be too risky for the private sector to undertake. Of particular importance are pilot inventions, the early proto- types of entirely new kinds of technology (Huffman and Evenson, 1993). For example, many crop breeding methods were developed at SAESs; private sector research on crop breeding became significant only after the introduction of hy- brid seed technology provided a natural means for a firm to protect its invest- ment, because producers were required to purchase new seed each year (Fuglie et al., 1996). In the pesticide industry, as well, innovations have come mainly from the private sector. Chemical companies do share information and conform to stan- dards on toxicities, dangers, and other factors, and there is a private/public part- nership in education and training for pesticide use. The key to cooperation and investment by the private sector has been regulation, combined with public re- search and extension programs which have protected producers and the environ- ment from misuse of pesticides. Markets for new technologies may sometimes appear to be too small to per- mit private firms to recoup their research and development costs, even if the invention would benefit society as a whole. In such cases, investment in private research and development will be lacking. In the long term, most of the benefits of technological improvements in agriculture accrue to consumers; competition eventually limits both producers and equipment suppliers to returns on invest- ment equal to the cost of capital. Traditional or conventional technologies may have fully captured the mar- ket, making it difficult for new technologies to emerge until some event occurs to disrupt markets. The long, slow development of conservation tillage illustrates the difficulty of penetrating a market that is dominated by conventional methods. Energy shocks in the mid- and late-1970s caused a disruption that conservation tillage, because of its resulting energy savings, could exploit, and conservation compliance policies added a further incentive to change from traditional tillage. Available producer and hired labor may not have sufficient training to use new technologies, limiting market potential. Lack of training in GIS and GPS electronics may limit adoption of these technologies and retard investment in them by the private sector. Some investments in making the systems easier to use, understand, and interface with computer and other systems may overcome an initial lack of training in the labor force. Finally, the private sector similarly has little incentive to engage in research aimed at enhancing environmental quality (i.e., by reducing pollutant emissions), even though the results of such research may be of great value to society. Because

PUBLIC POLICY AND PRECISION AGRICULTURE 95 agricultural pollutants are generally not regulated, and because environmental effects are by-products of production that do not show up on the bottom line, producers may be unwilling to pay for products embodying such research. Tech- nology development firms may thus have little or no incentive to engage in re- search to reduce environmental spillovers from agricultural activity. Increasing public concern for the environment may encourage technology providers and pro- ducers to adopt practices that enhance environmental quality, especially if they add little to production costs, but current legal or administrative requirements offer few direct incentives to do so (Fuglie et al., 1996). The allocation of research and development spending suggests that the pub- lic sector has largely concentrated on the areas where market incentives fail to generate private sector research interest. In 1992 research on plant and animal production systems accounted for 34 percent and 24 percent, respectively, of SAES research and development spending, and environment and natural resources accounted for an additional 24 percent. Relative shares of spending on these re- search areas at SAESs have remained largely unchanged for the past 20 years. By contrast, private sector agricultural research and development in 1992 was con- centrated on agricultural chemicals (37 percent) and development of end-use prod- ucts (30 percent) ( Fuglie et al., 1996). A relatively new form of research collaboration is the Cooperative Research and Development Agreement (CRADA), authorized by the 1980 Stevenson- Wydler Technology Innovation Act and its 1986 amendment, the Federal Tech- nology Transfer Act. This legislation permits federal laboratories to enter into agreements with universities, private companies, non-federal government enti- ties, and others to link the laboratory’s fundamental or pretechnology research capacity with the commercial research and marketing expertise of the private sector. The acts establish funding guidelines and rules regarding ownership of the intellectual property developed under CRADAs. Between 1987 and 1995, USDA entered into over 500 CRADAs, of which 227 remained active in 1995 (Fuglie et al., 1996). These agreements covered more than $61 million in research assets and resulted in 399 USDA patents generating $1.6 million in royalties in 1995. CRADAs were developed to increase the success of federal laboratories in intro- ducing new enabling technologies to potential uses in the private sector (National Research Council, 1995a). However, the National Research Council’s Committee on Criteria for Fed- eral Support for Research concluded that government resources supporting CRADAs could, in many cases, be better spent on other federal research initia- tives. They based this conclusion on recent criticisms of CRADA effectiveness, difficulty in analyzing CRADA effectiveness due to data inadequacy, uncertain- ties over ownership of intellectual property, and the small number of new jobs created (National Research Council, 1995a). CRADA research often replaces re- search that private firms would undertake in the absence of governmental agree- ments and may be particularly problematic in situations where there are many

96 PRECISION AGRICULTURE IN THE 21ST CENTURY start-up firms that could be potential CRADA collaborators. This situation char- acterizes many research areas in precision agriculture. Using CRADAs for preci- sion agriculture research and development would blur the distinctions between basic and applied research that this committee propose as criteria for appropriate public and private research roles relating to precision agriculture technologies. Specifically with regard to agronomic and crop management topics, the commit- tee concludes that the most valuable contributions public laboratories can make to precision agriculture are likely to be basic, pretechnology, nonappropriable research findings that can benefit all private sector developers of precision agri- culture technologies. These findings do not lend themselves to and should not be the subject of exclusive intellectual property agreements embodied in CRADAs. Priorities for public research and development in precision agriculture should be: • to invest in research areas in which improved understanding of variability is likely to make the greatest difference in terms of crop production meth- ods, farm profitability, or environmental quality; and • to invest in research areas likely to be neglected by the private sector, despite good prospects for significant benefits to society. While conceptually useful, practical distinctions between basic and applied research and fundamental technology development are increasingly blurred (Na- tional Research Council, 1995a). Nevertheless, this committee concurs with the National Research Council’s Committee on Criteria for Federal Support of Re- search and Development that the federal government should encourage, but not directly fund, private-sector technology development, except in direct pursuit of government missions and to develop new enabling technologies for which gov- ernment is the only available funder (National Research Council, 1995a). Their reasoning for this recommendation recognizes, as does this committee, that only where investments in research and technology cannot be fully captured by private sector firms is a prominent government role justified, particularly at the more applied end of the research spectrum. Development related to emerging technolo- gies, such as precision agriculture, may be an exception best dealt with by gov- ernment/industry partnerships, such as the Sematech industry consortium devel- oped to pursue research in semiconductor manufacturing technology. NEED FOR IMPROVED MEASUREMENT METHODS The potential of precision agriculture is limited by the lack of appropriate measurement and analysis techniques for agronomically important factors. Public sector support is needed for the advancement of data acquisition and analysis methods, including sensing technologies, sampling methods, data- base systems, and geospatial methods.

PUBLIC POLICY AND PRECISION AGRICULTURE 97 Although many of the technologies making up precision agriculture are rela- tively mature (i.e., GPS, GIS, and remote sensing), there remains room for im- provement in many technological areas directly related to agricultural applica- tions. One of the most important of these is the development of local sensors that can be used on farm equipment to determine crop stage, soil conditions and chem- istry, weed concentrations, presence of insects, and other variables important for crop growth. Public sector researchers should concentrate on the basic scientific principles that could underlie new sensor development and on the relationships between measurements from such new sensors and modeling of crop growth and yield. Private sector research and development is more appropriate for making the new sensors operational and marketable. Another farm management problem of special importance is the determina- tion of optimal sampling strategies. Some precision agriculture technologies func- tion by permitting adjustment of farming practices (i.e., input application rates) to match variability in production conditions, such as soil nutrient levels or other aspects of soil quality. Determining the extent of variability is essential, not only at the subfield level, but at all spatial levels. Optimal sampling depends on trade- offs between potential savings in input expenditures, potential gains from in- creased yields due to improved management, and sampling costs (Hennessy et al., 1996). Database and GIS systems include interpolation algorithms to predict data at intermediate points, but no existing research validates assumed projections under true agronomic variability in the presence of obvious measurement errors. Geospatial methods must be advanced and incorporated into GIS to facilitate accurate analysis and inference from collected precision agriculture data. The public sector should take the lead role in researching (a) the nature of variability within farm fields and at other spatial scales, (b) the required precision of com- patible measurements that are to be included within GIS data sets, and (c) the fundamental geospatial analysis methods necessary for interlayer correlation analysis and inference. Both the public and private sector have been involved in developing and disseminating standards for hardware, software, and data interpretation that could influence precision agriculture development and adoption. Such standards have been critical in the development of general computer technology (such as the ANSI, ASCII, and ISO 8211 data standards), and are emerging in GIS (such as the spatial data transfer standard and open GIS standards). Developing standards always involves a trade-off between ordering the chaos of individual systems and stifling creative breakthroughs in emerging technologies, and thus must be care- fully managed. From the perspective of the user, standardization would facilitate data inter- change, particularly moving spatial data from one proprietary software package to another and to regional databases. Hardware interoperability would facilitate connection of technologies and equipment into a unified system (i.e., a VRT con-

98 PRECISION AGRICULTURE IN THE 21ST CENTURY troller in a tractor cab communicating with separate rate controllers for seeds, liquid, and dry chemicals). Standards affecting data and hardware interchange affect the integration and ease of using these new technologies. Precision agricul- ture is technically possible today, in large measure, but requires a high degree of technical know-how and persistence, much as did the early personal computer systems. Precision agriculture developers and vendors are torn between conflicting goals of responding to user needs and maintaining proprietary advantages, market niches, and demand for system and electronic consulting services. There are also some potential conflicts between publicly provided services and private vendors. Most concerns in precision agriculture relate to spatial data standards, be- cause many aspects of conventional database management and operating systems already have information technology standards developed by industry and public consortia. These include: • government standards (i.e., the Federal Geographic Data Committee’s Spatial Data Transfer Standard and ISO 8211), • consortium standards (i.e., the Open GIS Foundation and the Agriculture Electronics Association [AEA]), and • ad hoc or default standards (i.e., from dominance in the market, such as AutoCad DXF or Arc/Info Export). Several paths of development and implementation could take place, with differ- ent trade-offs in timeliness, responsiveness, and enforceability. In the arena of precision farming, the Agricultural Electronics Association was founded by the Equipment Manufacturers Institute in 1995 to bring together diverse interests in the field of electronics in agriculture. Membership has grown from 19 original members to over 100 companies, organizations, users, and uni- versity and government liaison members. Subdivisions within AEA include a User Council, Equipment Council, Hardware Council, and Software/Information Systems Council. AEA identifies, develops, and facilitates appropriate action to increase compatibility and interchangeability of electronics and information sys- tems used in agriculture. AEA has made significant strides in promoting stan- dardization, including addressing issues such as • the interface between electronic equipment and specific connector, data, power and protocol requirements; • compatibility of electronics and information systems with precision farm- ing software; • a standardized data “reader” interface between chemical labels and ma- chines; • environmental standards; • livestock issues; • development of a database standard;

PUBLIC POLICY AND PRECISION AGRICULTURE 99 • a common communication structure; • standardized static and dynamic spatial data exchange formats; and • data dictionary specifications in ISO 8211 format for yield (grain crops), soil fertility, and application crop plan characteristics. NEED FOR UNBIASED EVALUATION Unbiased, systematic, rigorous evaluations of the economic and environ- mental benefits and costs of precision agricultural methods are needed. USDA should facilitate and coordinate evaluations conducted through col- laborations of public agencies, professional organizations, commercial or- ganizations, and producers. An appropriate role for public agencies is independent, objective evaluation of precision agriculture technologies. Private technology development firms and input suppliers have a natural commercial interest in promoting precision agricul- ture. Individual producers may have insufficient incentives or resources to con- duct evaluations or make the results known because all producers in a region can learn from those experiences at little cost, creating a “free rider” problem. The benefits of having local information about the performance of precision agricul- ture technologies exceed those a single producer can gain, arguing for a public role. Moreover, producers may find it difficult to apply the experiences of a single farm to their own situations because they may not be able to make the appropriate adjustments for differences in conditions across sites. Site-specific factors can be so important in evaluating these technologies that the usual producer network will likely be inadequate for disseminating precision agriculture information. Producers need unbiased assessments of precision agriculture’s performance characteristics under various conditions. Public and private environmental orga- nizations are also interested in unbiased evaluations of precision agriculture’s environmental performance (Ogg, 1995). Acceptance and support for precision agriculture depends on the extent to which potential efficiency gains and environ- mental benefits are actually achieved. USDA is in a unique position to facilitate and coordinate evaluation and research activities among federal agencies. USDA and its affiliated SAES part- ners have the agronomic knowledge necessary to evaluate the effectiveness of specific precision agriculture technologies and systems. Where federal agencies outside agriculture have some basic technological components and expertise nec- essary to advance precision agriculture, collaboration in that evaluation should be encouraged. Producers and other customers for precision agriculture technologies should be encouraged to search for multiple sources of information when deciding whether to adopt particular components of precision agriculture technology. Pro- ducer decision-making processes are complex, and multiple sources of informa-

100 PRECISION AGRICULTURE IN THE 21ST CENTURY tion would help to shape and confirm decisions. Evaluation does not necessarily imply that producers are not capable of making sound decisions or that input suppliers are untruthful in their claims. Research in the area of decision theory indicates that sound judgments are particularly difficult for humans in situations where there is considerable variability, where there are time lags between actions and results, and where there are multiple and complex cause-and-effect relation- ships. These three characteristics seem to apply to precision agriculture and its adoption and evaluation. Several factors will make it difficult for public agencies to carry out such evaluations in the area of precision agriculture. The technology is changing so rapidly that evaluations of specific technology components have a very short useful life. Producers will not be well-served if equipment or products embody- ing technologies which have been evaluated cannot be compared with newer, more sophisticated versions that have not been evaluated. Products that have been validated in the field before submission for evaluation, whether prototypes by industry or federal expertise, should be compared on the same basis as existing commercial products. System evaluations are appropriate on technologies that are installed, maintained, and operated as specified by the manufacturer. Because the area is evolving so rapidly, technology developers may be reluctant to expose newly developed technology to public evaluation, risking loss of proprietary and trade secret information. Meaningful collaboration between private firms and public agencies, and between agencies, may not be forthcoming without consid- erable effort. Over the long term, there is no substitute for carefully designed observation of economic and environmental results obtained by actual producers in real field conditions. For these experiments to be useful, side-by-side treatments and statis- tical control methods need to be used to distinguish precision agriculture tech- nology’s contributions from normal variation in resources, weather, and manage- ment. Given the systems nature of precision agriculture techniques and the importance of site-specific variability, on-farm experimentation performed in collaboration with producers will be necessary and desirable, compared with more traditional farms of plot-based research design (Alliance on Agricultural Infor- mation Technology, 1996). Research collaborators can mine a wealth of on-farm data and use regression analyses and other multivariate statistical methods to isolate the multiple sources of variations that influence economic and environ- mental outcomes of precision agriculture. These findings can provide invaluable guidance to producers on the expected benefits from adoption of precision agri- culture technologies in their particular setting. The accuracy and reliability of methods for collecting precision data need to be evaluated to ensure confidence in grid soil-sampling schemes, directed sam- pling, and yield monitor results (Blackmer and Schepers, 1996; Lamb et al., 1995). Similarly, the accuracy and reliability of methods for making precision applica- tions of fertilizers, pesticides, irrigation water, and other inputs also need confir-

PUBLIC POLICY AND PRECISION AGRICULTURE 101 mation (Chaplin et al., 1995; Olieslagers et al., 1995). Finally, producers need confirmation that new methods of interpreting precision input information to de- velop recommendations for management changes are accurate and do improve economic and environmental performance over whole-field management meth- ods (Booltink et al., 1996; Rawlins, 1996). Subfield fertilizer response relation- ships, economic pest control thresholds, and computer-based decision support systems derived from crop growth simulations need to be tested and evaluated under actual field conditions in a variety of circumstances (Heiniger, 1996). Precision agriculture evaluation activities should be undertaken by both the public and private sectors. Organizations in both sectors should work together to avoid possible biases in evaluating the efficacy of the technologies. Because of the site-specific nature of the farm fields that precision agricul- ture is designed to address, evaluation cannot be generalized and must be couched in terms of the specific resource conditions to which it is applied. Evaluations should compare precision agriculture systems with conventional, uniform man- agement systems, and against each other, recognizing that precision agriculture enables changes beyond variable-rate application of inputs. Evaluation is part of an iterative continuous cycle of research, development, and deployment that is necessary for the technologies to evolve and improve. NEED FOR NEW APPROACHES TO RESEARCH Precision agriculture requires new approaches to research that are designed explicitly to improve understanding of the complex interactions between multiple factors affecting crop growth and farm decision making. USDA and land grant universities should give increased priority to such new approaches by reallocating personnel and budgets. The most important research area of precision agriculture is development of theoretical and empirical knowledge to support improved crop models, farm man- agement methods, and expert systems software. Much of the discussion of preci- sion farming has revolved around measurement of variability through yield moni- tors, remote sensing, and digitized soil mapping. However, measurement means little if it does not result in better management. In this regard, precision agricul- ture is a systems approach to agricultural management, not dissimilar to the appli- cation of systems principles in other arenas since the 1950s, including environ- mental problems in the 1970s, IPM and sustainable agricultural systems in the 1980s, and watershed and ecosystem management initiatives in the 1990s. Preci- sion agriculture is fundamentally an information technology that focuses people on the complexities and interrelatedness of agroecosystems in a holistic way. If systematic understanding of the cropping system is not captured in models, how- ever, it is unlikely that the volumes of precision data on subfield variation can be meaningfully processed to provide improved management decisions. If such mod-

102 PRECISION AGRICULTURE IN THE 21ST CENTURY els cannot be developed, much of the potential of precision agriculture will not be realized. Reliable crop models are the foundation of any attempt to construct data- driven, computer-based decision support systems that can effectively use preci- sion data to make precision management changes, yet few such models exist. Theoretical and empirical understanding of crop yield responses to variations in nutrients and soil quality remains primitive. For example, fertilizer recommenda- tions are based on rules of thumb, such as “one and a quarter pounds of nitrogen per acre for every bushel of yield goal for corn,” even though theory and evidence indicate that crop nutrient response is nonlinear and that yield will not respond to additions of some nutrients when others are limiting (Cerrato and Blackmer, 1990; Chambers and Lichtenberg, 1996; Frank et al., 1990; Paris, 1992; Paris and Knapp, 1989). Similarly, it is well-known in principle that crop productivity and crop nutrient response depend on elements of soil quality and tilth (such as struc- ture, texture, organic matter, and water-holding capacity), yet there exists little quantitative modeling of these factors in crop production. Pest management pre- sents a similar picture; treatment thresholds are frequently based on rules of thumb because there are no reliable crop-pest ecosystem models. Even if rules of thumb for nutrient and pest management are based on significant experimental data, the resulting recommendations are developed on an aggregated basis that ignores the other factors that vary within fields. “[Precision agriculture] departs from current nitrogen fertilizer guidelines that were primarily developed on a regional scale. . . . As a result, current nitrogen recommendations may have limited application to site-specific nitrogen management.” (Pan et al., 1997, p. 81). If modeled relation- ships can be developed that capture the effects of variation in factors that vary at the subfield level, better recommendations can be made. Nutrients, pest management, and soil quality are obvious targets for public research because of their linkage to environmental quality. Nutrient pollution of both surface water and groundwater is a significant problem throughout the United States, and agriculture is a major contributor to nutrient pollution in many areas, such as the Great Lakes region, the Chesapeake Bay watershed, and the Missis- sippi drainage (where it contributes to Gulf of Mexico hypoxic zone problems). In many areas, reductions in nitrogen pollution resulting from improved nitrogen management could justify the use of variable-rate application and other precision farming methods even when reductions in fertilizer expenditures do not (Khanna and Zilberman, 1996). Public incentives, such as regulations, cost-sharing, or incentive payments, may be needed to spur adoption in such areas, if it is not otherwise profitable. Models of crop-pest interactions are important for devising improved ecologically-based pest management strategies. More judicious use of pesticides could reduce environmental damage. Investment in soil quality is a central tenet of sustainable farming systems. Models elucidating the relative con- tributions of the components of soil quality to crop yield could help improve the design of sustainable farming systems.

PUBLIC POLICY AND PRECISION AGRICULTURE 103 Development of such crop models will involve basic research into the effect of subfield variation in availability of nutrients for crop growth, the effect of variation in soil quality on crop nutrient uptake and pest (insect, weed, and dis- ease) prevalence, and crop-pest ecosystem interactions. It will also involve ap- plied empirical research attempting to quantify crop yield responses to nutrients, soil quality, and pest prevalence across the growing season for important crops in different locations. It will require interdisciplinary study involving agronomists, entomologists, economists, plant pathologists, weed scientists, and ecologists. Developing such models will not be an easy undertaking, and efforts requiring coordination across many disciplines are essential. The current research model for crop sciences employs carefully controlled research plot experiments with replicated block designs in which only the factor of interest (i.e., fertilizer rate) is allowed to vary. Because precision agriculture has already been identified as a technology of some promise, detailed small-plot studies and technology-adaptation experiments may not be necessary (Gomez and Gomez, 1984; Gotway Crawford et al., 1997). Further, the system may need to evolve so that innovation and learning can exploit both traditional research plot experiments and information captured from actual field operations through preci- sion agriculture. Precision agriculture has the potential to collect many layers of data for entire fields and record detailed variation in many variables that affect crop growth. Thus, precision agriculture could change the research paradigm from station-based plot studies to farm-based studies by (a) using more complex ex- perimental designs, such as incomplete block designs, row-column designs, near- est neighbor designs, and split plot designs; (b) specifically incorporating spatial variability in experimental designs; (c) supplementing mean-based analyses with comparisons of entire distributions; and (d) using statistical methods such as multiple regression (Gotway Crawford et al., 1997). Under this paradigm, groups of producers would collect precision data; agronomic researchers would analyze the data with statistical methods to estimate how small changes in manageable factors affect crop yield in various resource and weather situations. The incen- tives for and obstacles to producer data sharing need to be fully explored and carefully understood if such cooperative on-farm research is to succeed (see sec- tions on data ownership and privacy). Improved farm management methods are equally necessary. At present, farm management models are based on budgeting, which assumes fixed input-output ratios. Recommendations based on improved crop models will likely need to be derived from nonlinear optimization, such as profit maximization, because such crop models will likely be characterized by variable input-output ratios which may change throughout the growing season. Stochastic optimization frameworks may be needed to take into account the random occurrence of rainfall. Dynamic optimization may also be needed to take into account changes in recommenda- tions as the growing season progresses. As discussed previously, it is not clear a priori that precision agriculture

104 PRECISION AGRICULTURE IN THE 21ST CENTURY technologies will reduce environmental spillovers from agriculture. For example, if increased precision in application rates increases crop yields sufficiently, then producers may have an incentive to increase rather than decrease application rates. Thus, research explicitly aimed at elucidating the environmental effects of preci- sion agriculture technologies should also be a priority (Larson et al., 1997). Decision support system models force consideration of factors and interac- tions affecting the entire system. Therefore, work oriented toward developing such models requires institutional arrangements that cut across strict disciplinary lines. The relationships represented in such models are also inherently multivari- ate, lending themselves more toward on-farm research than plot-based studies. Reorienting research priorities toward model development and validation could thus alter institutional incentives for more holistic, farm-based, decision-oriented research efforts. TRAINING AND EDUCATION NEEDS In the twenty-first century agricultural professionals using information tech- nologies will play an increasingly important role in crop production and natural resource management. It is imperative that educational institutions modify their curricula and teaching methods to educate and train students and professionals in the interdisciplinary approaches underlying precision agriculture. For use of precision agriculture to become widespread, producers and pro- spective employees will need general computing skills and technical literacy. Specific skills needed by service specialists, such as high-tech equipment opera- tors and GIS or GPS technicians, could be taught both through traditional four- year programs and vocational training. Consultants, system integrators, and oth- ers with an understanding of how to develop and apply precision agriculture will likely need postgraduate education. Successful training of users of information technologies will require disciplinary depth (i.e., agronomy, agricultural engi- neering, and soils) and analytic skills (i.e., spatial analysis and crop modeling), that is best provided by long-term education with emphasis on interdisciplinary synthesis. The mind-set that is needed to ensure the beneficial use of precision agriculture should be fostered in educational institutions, particularly in elements of programs that provide an understanding of technologies in a broader context. While some institutions have developed undergraduate courses and extension education programs in precision agriculture technologies, a more systematic re- appraisal of their programs is needed. The content and form of agricultural education and extension are evolving under a number of pressures. A recent National Research Council report docu- mented problems with the current land grant system and provided recommenda- tions for sweeping reforms (National Research Council, 1996a). Many of these

PUBLIC POLICY AND PRECISION AGRICULTURE 105 were intended to provide better accountability to the broader social purposes for maintaining these institutions. Significant impetus for change comes from the declining public support for public education institutions in state and federal bud- gets. Increasing use of several kinds of distance education and on-the-farm re- search are examples of how the form of education and extension are evolving. As part of the transition of parts of agriculture from agrarian to industrial processes, new domains such as precision agriculture are exerting a demand for new services. In this case, it is a demand for help in understanding whether and how to use the technologies and a demand for skilled graduates to design, imple- ment, and operate the systems. Some of these training and education gaps are being filled by the private sector. A significant challenge faced by public educa- tion institutions is to evolve a role that allows them to serve a broader social purpose, to guide the technology and turn out graduates who understand preci- sion agriculture in a broader socioeconomic and environmental context, while still providing useful practical skills to graduates and useful information to users of the technology. Using systematic, holistic, information-driven production management as an organizing principle for agricultural education shifts the focus to encompass both the broader educational goals and the details of precision agriculture systems within a new framework. This synthesis could evolve from replacing narrow dis- ciplinary frameworks with the concept of systems derived from crop modeling work (Stone, 1989). Given the development of precision agricultural technologies outside tradi- tional agricultural institutions, it is not clear who should provide scientific, tech- nical, and managerial education needed for precision agriculture. In many cases, this expertise is provided by several university departments. Although it is un- likely that computer, geography, and engineering departments will begin teach- ing agronomy and other agricultural sciences necessary for precision agriculture, agricultural departments should seriously consider recruiting students with back- grounds in computer, geography, and engineering technologies that are driving precision agriculture. It needs to be determined whether, in the short run, it is more efficient to teach basics of agricultural sciences to students already familiar with the technologies used in precision agriculture or to do the reverse. Designers of continuing education curricula will perhaps have an easier task. Material will have to be developed for both technologically sophisticated students needing more agricultural science background and agriculturists needing more training in the technologies. The wisest course may be to overcome institutional barriers on cam- pus and develop creative syntheses between computer, geography, and engineer- ing departments and agricultural science departments for courses that draw on the strengths of each (see for example, National Research Council, 1997). Similar to the demand being placed on research in public institutions, exten- sion agents, consultants, suppliers, and others who “retail” information to farmers will have corresponding requests to provide help with understanding and using

106 PRECISION AGRICULTURE IN THE 21ST CENTURY the technologies. Professional organizations involved in precision agriculture, such as the Alliance on Agricultural Information Technology, have already called for “clearinghouses,” unbiased sources of comparative information on alternative precision agriculture methods which could be set up and operated by extension services or others who work directly with producers. A broader public purpose would be served if the clearinghouse function included unbiased comparisons of benefits, costs, and effects of precision agriculture adoption and use. Information service providers could help address specific problems that arise from precision agriculture and could play a role in promoting socially beneficial aspects, such as environmentally sound approaches. In the legal arena, templates, forms, and model contracts are needed to avoid conflicts associated with data ownership, intellectual property rights, and the protection of privacy. The extension system would seem to be suited to such a public information role, but deficiencies in specialized scientific and technical training of current agents raises questions of capacity and capability. The rapid technological changes occurring in diverse areas of precision agriculture, including chemical pest management, environmental improvement, and farm electronics, present an enormous challenge to extension. Moreover, the need to integrate these technolo- gies within a systems view of agricultural management makes its leadership in precision agriculture information problematic. NEED FOR HIGH-SPEED CONNECTIVITY High-speed data connectivity is needed in rural areas to support precision agriculture. Agricultural organizations and agencies should work collab- oratively with public agencies and industries to ensure adequate rural con- nectivity. The communications infrastructure such as satellites, high-speed telecom- munications services, and the Internet will be essential if precision agriculture is to develop to its full potential. Extensive adoption of precision agriculture will depend on access to a modern information infrastructure in rural areas, particu- larly telecommunication services such as the Internet. Such communications ser- vices will similarly be essential for taking full advantage of the data-generation capabilities of precision agriculture technologies described previously. Private industry may provide little of this communications infrastructure. Infrastructure investments are characterized by substantial economies of scale because the cost of building the infrastructure is high, whereas the cost of provid- ing service to additional customers once the infrastructure is in place tends to be low. Firms can recoup the needed investment in infrastructure by charging suffi- ciently high access fees and usage charges. Such charges will tend to inhibit use, however, so that use by some firms and individuals will remain below efficient levels. Moreover, private industry may not find it profitable to provide infrastruc-

PUBLIC POLICY AND PRECISION AGRICULTURE 107 ture coverage to all of the areas of the country. In this case, benefits to society as a whole can exceed the investment costs. Rural areas may be particularly at risk because customer density is low. Development of the communications infrastructure may be an important stimulus for rural economic development. Congress requested that the General Accounting Office study how Internet and high-speed telecommunications ac- cess for rural areas could be promoted for economic development purposes through new and existing federal programs (General Accounting Office, 1996). The Telecommunications Act of 1996, although not specifically addressing pre- cision agriculture, regulates the competitive framework within which informa- tion technology services will be provided in rural areas. Within that framework, state telecommunications regulatory agencies, state and local governments and other rural institutions, such as university extension departments, will have an interest in encouraging rural access to advanced telecommunications. Access to skilled labor and up-to-date information about technical processes and market conditions have been important reasons for concentrating industries in certain locations in the past, leading to faster growth in urban than in rural areas (Krugman, 1993). Modern communications are reducing the advantages of geo- graphic concentration. If competitors of the United States in global food and fiber markets adopt precision agriculture technology more readily than do domestic producers and are perceived to gain an economic advantage, public policies encouraging preci- sion agriculture adoption may be developed to maintain global competitiveness. The President’s Council on Sustainable Development found precision agriculture to be one of the environmental technologies that offer potential for domestic eco- nomic development through increases to domestic productivity and development of export markets in environmental technology (Sustainable Agriculture Task Force, 1996). Although the notion of universal access to telecommunications has been embodied in federal policy documents such as Principles for a National Informa- tion Infrastructure, it is unclear whether and how universal access will become a reality (U.S. Information Infrastructure Task Force, 1995). Some experts suggest that the competition created by the Telecommunications Act of 1996 will be suf- ficient; others believe federal subsidies comparable to the rural electrification programs will be needed. In either case, access to information services will affect rural areas. To the extent that it takes place in different forms and at different rates of implementation, equity issues will arise. Advocacy groups are already concerned about disenfranchisement of information-poor socioeconomic groups. This potential exists in rural areas, particularly areas without activities such as precision agriculture providing an impetus for commercial enterprises to provide services. Fast, reliable Internet access in remote rural areas will affect precision agri- culture approaches that rely on regional aggregation and distribution of data or

108 PRECISION AGRICULTURE IN THE 21ST CENTURY communication with experts, such as crop monitoring or interpretations of field data. The Alliance on Agricultural Information Technology identified rural band- width, GPS differential correction, and digital orthophotography as key informa- tion technology components of precision agriculture that currently restrict its adoption (Alliance on Agricultural Information Technology, 1996). This infor- mation infrastructure is a significant focus of a Clinton administration initiative, the National Information Infrastructure Council and the related National Infor- mation Infrastructure Task Force. Although not specific to precision agriculture, these groups have provided a set of principles and recommendations that would result in significant improvement in data access for rural areas. The components of an information infrastructure for precision agriculture include access to spatial technologies such as GPS and remotely sensed imagery, which will spill over into areas such as modernization of land records and land use planning. In much of the rural United States, local government land informa- tion systems are antiquated and unable to provide timely, reliable information for private or government land-related decision making. In some areas, precision agriculture will be a useful force in improving these systems, in addition to pro- viding some of the automated data and spatial technologies that will be required. CLARIFICATION OF INTELLECTUAL PROPERTY RIGHTS, DATA OWNERSHIP, AND DATA PRIVACY Precision agriculture will require clarification of intellectual property, data ownership, and data privacy rights. The extension service should play a lead- ership role in providing education on existing law pertaining to these issues. Intellectual property rights and data ownership are evolving areas of concern in terms of both information technologies and legislative and judicial activity. Precision agriculture raises some unique questions relative to data ownership be- cause of the spatially extensive nature of the resources involved and because of its related information gathering systems, including remote sensing and third- party data providers. Despite the continuing evolution of intellectual property law, legal precedents from the computer industry and general business practice provide guidelines applicable to precision agriculture. Legal complications need not constrain adoption of precision agriculture technologies if the legal forms from other industries (copyright, trade secrets, and patents) can be translated to precision agriculture, allowing producers to abandon handshake agreements and formalize their legal rights to their data. Both the American Farm Bureau Federa- tion and the Agricultural Electronics Association have been developing legal tem- plates and forms for producers to use in asserting ownership over precision agri- culture data. The extension service or legal experts associated with SAES could provide a valuable service by working with these and other groups, adapting the

PUBLIC POLICY AND PRECISION AGRICULTURE 109 model legal forms for data disclosure relating to precision agriculture data, and ensuring that they receive widespread dissemination and adoption. Data ownership issues could affect the adoption and value of precision agri- culture. A balance between protections for individual data ownership and ben- efits to multiple users must be found. Two scenarios could result: • If ownership cannot be or is not protected, there may be a chilling effect on the willingness of individuals to provide field and farm data to aggre- gate databases, whether publicly or privately established. • If ownership protections require producers to jealously guard their indi- vidual data, broadening the value of individual data collection through regional aggregation for area-wide crop management research or recom- mendations will be retarded and made prohibitively expensive. Providers of information services, such as fertilizer dealers providing preci- sion application services, generally recognize the land owner or farm manager as the de jure owner of data. This probably is supported by law, because the pro- ducer is buying the information services from the provider. Ownership issues may be further complicated by contractual arrangements between producers and providers, data acquisitions or exchanges by providers or third parties, and at- tempts to copyright data compilations. However, mere possession of the data by service providers, supported by the provider’s intellectual investment in storing and analyzing the data, may lead to appropriation of data rights, unless the operator’s rights are specified with appropriate legal documents. Although these concerns are novel for agriculture, many of the same issues have been faced by other industries. Some tools, such as copyright, cannot protect raw data but can protect the expression of ideas or concepts embodied in the data, such as a set of recommendations, a computer model, or a compilation of the data. Trade secret protection does not apply to raw data unless the producer can show that the data he collects, or pays to have collected for him, meet the specific criteria for trade secrets. The data must derive independent actual or potential economic value from not being generally known to other persons who could profit by it, and the producer must show that reasonable efforts were made to maintain that secrecy, such as a nondisclosure agreement, a license agreement, or some other legal instrument that restricts access and disclosure to others. Once such legal forms deriving from other industries are adapted to the peculiarities of pre- cision agriculture, many of the issues of data ownership will be resolved. Because of the proprietary nature of computerized systems, the producer may get data in a format useable only by the provider, who may then exercise de facto control over the data. These are not legal ownership issues, but issues of technol- ogy and technological competence. The producer’s recourse, in this example, may be the expensive one of paying for data conversion to the nonproprietary format so another provider can work with the data. This may be financially oner- ous but it presents no legal barriers to clear data ownership.

110 PRECISION AGRICULTURE IN THE 21ST CENTURY The value of individual field and farm data increases when it is collected across a region and integrated with data from other sources and other farms (see section on data assembly and aggregation, below). Aggregating data has implica- tions for ownership as well. Some farm organizations have asserted a right to create such collections, presumably to provide better information services to members and to develop regional strategies (American Farm Bureau Federation, 1995). Once the producer fails to assert ownership over precision data, compilations and abstractions from that data may be difficult to protect. If data are not handled as trade secrets, they become part of the public domain and proprietary rights are lost. Although raw data cannot be copyrighted, compilations of raw data that have been selected and arranged in creative ways can be protected through copy- right (Feist Publications, Inc. v. Rural Tele. Serv. Co. Inc., 499 U.S. 340 [1991]). Vendors with regional databases can protect content to the extent allowed by merger doctrine. “When the expression of an idea is inseparable from the idea itself, the expression and the idea merge . . . .” (Holland, 1994). That is, copyright can protect a unique way of managing, analyzing, or displaying the data that merges inseparably with the data and that cannot otherwise be copyrighted. Intellectual property rights in precision agriculture products and software are protected in the same way that other computer hardware and software are protected. Copyright and patent laws apply to these creations just as for creative products of other industries. In general, the more basic the scientific finding underlying a new development, the less protection these traditional property rights instruments afford. Patents and other forms of intellectual property rights to knowledge were created to mitigate the problem of underinvestment in re- search and development. Patents confer a temporary monopoly (17 years) on the fruits of research, allowing patent owners to recover the costs of research. As discussed above, intellectual property rights do not apply to pretechnology sci- ence (Huffman and Evenson, 1993). Another body of law governs privacy issues associated with ownership and use of data by governments relative to individuals. Traditional privacy issues such as personal information on health and income are protected from disclosure by statute. For example, responses to the Census of Agriculture are protected from disclosure and can be published only in aggregate form. Land information in the public record is generally considered part of public domain (i.e., land own- ership and real property taxes), including farm ownership and management infor- mation for federal conservation cross-compliance and many state programs. Some forms of remote sensing may already cross the threshold into invasion of privacy (depending on purpose, access, etc.). This may depend on the degree of intrusiveness or “subjective expectation of privacy” (Gabrynowicz, 1996). Regulatory access is often a test of the Fourth Amendment (illegal search and seizure). The boundaries between public and private collection and use (open records laws) may make it difficult to determine what is subject to the Freedom of

PUBLIC POLICY AND PRECISION AGRICULTURE 111 Information Act. A patchwork of laws, often poorly enforced, addresses restric- tions on public agency use and disclosure of individual data (Onsrud et al., 1994). In GIS, private sector operations, fearing loss of data or proprietary advantage, have been reluctant to participate in multipurpose land information systems where their data would be intermingled with other data in a public system (i.e., private utilities participating in local land records systems). By analogy, public-private cooperation in precision agriculture could be inhibited. Several groups advocated changes to current law to clarify intellectual property rights in databases. The National Information Infrastructure Task Force suggested several revisions to copyright law to incorporate changes related to information technologies without fundamentally changing the system (U.S. Information In- frastructure Task Force, 1995). The American Farm Bureau Federation (1995), in a white paper on information technologies, advocated statutory revision of the Copyright Act to protect databases developed from collections of farm- and field- specific information. The American Committee for Interoperable Systems argued that copyright should not be used to inhibit interoperability of operating systems and software across computer platforms (American Committee for Interoperable Systems, 1994). Producers have expressed reservation that precision agriculture data may be used by government agencies for regulatory purposes. These new sources of in- formation, however, will have the same privacy protections against use by gov- ernment agencies as traditional sources of farm information, such as farm records, weigh bills, and other private documents. NEED FOR DATA ASSEMBLY AND AGGREGATION Data collected for use at the subfield and field levels have additional value for research, testing, evaluation, and marketing when assembled into re- gional databases. Mechanisms are needed to create these databases and make the data available for these additional uses including data collection and transfer standards; institutions for collecting, managing, or networking data; and policies to facilitate data sharing and access, while protecting proprietary interests and confidentiality. As valuable as precision agriculture data may prove to be to individual land- owners, much of the potential value of the huge amount of electronic data that could be collected by these technologies will not be realized unless the individual farm databases are consolidated into regional databases. These would not be av- erages or other statistical summaries of detailed data, but massive compilations of the detailed data itself, without information identifying individual farms from which the data are collected. Summaries might be made from these data for some purposes, but the detailed data needs to be accessible for analysis and modeling of relationships between inputs and outputs, including environmental outcomes;

112 PRECISION AGRICULTURE IN THE 21ST CENTURY the data also needs to be sufficient to control for other sources of variation. Analy- sis and modeling of crop responses to a variety of soil, weather, pest, and other crop management factors can best be done when the most complete range of variation is present in the data. An individual producer’s database is limited to the trials and management actions that form the recent history on his or her farm, whereas a regional database would encompass a far greater variety of manage- ment responses to similar conditions. A regional database might include condi- tions that did not occur on an individual farm one year but could occur in the following years. Researchers, extension agents, input suppliers, commodity com- panies, and government officials as well as producers will be interested in such regional databases if they can be made available. Despite the apparent benefits accruing from combining data into regional databases, it is by no means certain or inevitable that such data will be assembled. Information service providers that currently collect or are given precision agri- cultural data have proprietary interests in restricting access to their customer base, or in getting remuneration for access. In the absence of legal safeguards for data privacy, producers may be reluctant to share data on their operations in a freely accessible, voluntary regional database. Public agencies may not have the organi- zation, knowledge, or resources to develop regional data sharing cooperatives that could allow effective use of such data. Two kinds of obstacles stand in the way of creating regional databases from farm and field microdata. The first, and more surmountable, are technical barriers such as computer capacity (whether centralized or distributed as a network) and transfer standards and protocols. Second are institutional barriers, such as lack of clear leadership roles in establishing the databases, legal issues of data ownership and privacy discussed above, and issues of compensation and access. Many of these issues were discussed in relation to recommendations above and will not be repeated here. Although many questions remain to be answered, data aggregation is already occurring, particularly in areas where precision agriculture is implemented by input suppliers that are, by default, collecting regional databases. There is thus some urgency to resolve outstanding issues before problems with existing data aggregations surface. NEED FOR REVIEW OF PUBLIC DATA COLLECTION The methods and purposes of publicly funded data collection activities should be periodically reviewed and adjusted to ensure that data are accessible and useful for precision agriculture as well as supportive of other public and private purposes. The National Cooperative Soil Survey should revise exist- ing procedures to make more effective use of information technologies, farm- generated data, and new concepts in soil science.

PUBLIC POLICY AND PRECISION AGRICULTURE 113 Public sector investment in data collection and management is often driven by legislative mandates or specific operational missions. As the ability to collect, manage, and particularly share data improves with improvements in information technologies, and as budgets for public data collection decline, it becomes even more important to gather data that balance specific agency and program require- ments with broader purposes. While the current extent of precision agriculture adoption limits what can be accomplished, agencies need to carefully examine precision agriculture, both to ensure that the agencies are providing data useful to producers using precision methods, if that is appropriate, and to assess the poten- tial for using data collected on the farm to supplement or replace existing data collection efforts. Finely detailed information about soil properties is fundamental to some types of precision agriculture. To generate such data, producers or consultants have used various strategies for fine-scale soil sampling, including grid sampling at various spatial frequencies and sampling schemes keyed to landscape character- istics, such as topography and drainage. In other forms of precision agriculture, the data may come from sensors, such as yield monitors, on-the-go sensors, or aerial photography. These applications may still benefit from detailed soils infor- mation available in a form that can be integrated with other digital data. For soils and other kinds of agricultural data that federal agencies have traditionally col- lected, widespread adoption of precision agriculture should motivate review of existing efforts and exploration of new opportunities in using precise data gath- ered on farms. As an example of how precision agriculture has the potential to both change what data products are provided and how data is collected, we examine the Na- tional Cooperative Soil Survey (NCSS). NCSS, a partnership of the Natural Re- sources Conservation Service (NRCS) with local and state agencies and land grant institutions, has been generating soils information for several decades. Although originally focused on supporting agricultural uses of soils data, the mission of NRCS and NCSS is now a much broader one, that of managing the nation’s soil resources and providing data and technical support “to help people conserve, improve, and sustain our natural resources and environment” (Natural Resources Conservation Service, 1995). The NCSS products are not useful for precision agriculture for several rea- sons. First, the published Soil Taxonomy and the methods in the NRCS Soil Survey Manual are focused on the pedon concept and soil classification. These publications tend to be oriented toward soil homogeneity, whereas precision agri- culture needs additional information about soil variability. Second, NCSS has had a goal of nationwide uniformity in their products, although the requirements of potential users vary widely. The result may be products that are compromises between the needs of many users but that do not completely suit any user. For example, the map scales chosen for detailed soil surveys (typically in the range of 1:12,000 to 1:24,000) are convenient scales for surveying and cartography but do

114 PRECISION AGRICULTURE IN THE 21ST CENTURY BOX 4-1 Federal Data Collection Efforts Most federal data collection efforts centering around agricultural pro- duction are derived from an extension model where enumerators or sci- entists from federal agencies collected relatively sparse data from farms, summarized and analyzed the data, and published findings for state or regional aggregates, sometimes by broad classes of farms. The pro- ducer’s role in this process was passive: responding to questions posed by agency personnel and receiving published reports, sometimes with assistance from extension personnel to see how the results applied to their producer’s particular farm. Precision agriculture data collection has the potential to revise this model in several important ways, because producers are now able to collect far more specific and detailed data more efficiently than federal agencies can. In this new paradigm, producers, or their consultants and suppliers, would collect the data on a precision basis, perhaps according to some standardized metadata protocol. The data would be gathered in centralized databases or data warehouses run by agencies, coopera- tives, industry groups, or private enterprises. Agencies may pay produc- ers for data collection or may pay intermediary data cooperatives or firms for access to the databases. Agencies may produce the same kinds of summary reports for the public as in the past, but may make available more specific and detailed analyses for individual producers, or may pro- vide detailed databases for producers and their advisors to use. The producer’s role in this system would be more active because the data collection would be designed primarily to serve the producer’s informa- tion needs and only secondarily to contribute to a larger database. The agency’s role would be less about deciding what questions to ask and more about investigating what can be learned from the available data. Some of the more prominent examples of federal agency data collection efforts that could be transformed in a world where precision agriculture is widely adopted are briefly explored below. NATIONAL AGRICULTURAL STATISTICS SERVICE The primary sources of information for the National Agricultural Statis- tics Service (NASS) are farmers and ranchers, livestock feeders, slaugh- terhouse managers, grain elevator operators, and other agribusiness personnel. NASS relies on survey respondents’ cooperation in voluntar- ily supplying data for the reports, and NASS holds confidential all data on individual operations. Objective yield surveys are conducted during the growing season to monitor crop conditions and yields in thousands of fields by enumerators who count the number of plants and, later in the season, count and measure ears, pods, bolls, and so on. The crop devel-

PUBLIC POLICY AND PRECISION AGRICULTURE 115 BOX 4-1 Continued opment data gathered through these objective yield surveys are used to forecast yields or project production (i.e., for wheat, cotton, soybean, potato, burley tobacco, onion, and a variety of fruit and nut crops). When the farmer harvests fields containing the plots, enumerators make their final visits to the sample plots to determine harvesting losses and esti- mate net yields. With new authority to conduct the Census of Agriculture, formerly in the Bureau of Census, NASS will collect information on the acreage in various farm uses, crops and livestock produced, and sales of agricultural products, as well as socioeconomic information on each farm operator and his or her family. Information gathered by satellites supplements that collected by enu- merators. Current satellite technology (LANDSAT and NOAA-AVHRR) applied to crop estimates has certain limitations; more frequent coverage is needed, and satellite scans can be rendered ineffective by cloud cover. Until commercial satellites overcome these restrictions, the NASS re- mote sensing program will remain limited. However, the data are excel- lent for timely views of large areas that are behind or ahead of previous seasons, or areas that are under stress caused by drought, excessive moisture, or disease. Widespread adoption of precision agriculture meth- ods could provide more detailed data from a larger number of producers, while integrating soil and weather data which could lead to greater under- standing of the causes of spatial and temporal variations in crop and livestock production. While the potential for such data collection is cur- rently limited, NASS should investigate possibilities for precision agricul- ture data to augment conventional data collection methods in the future. Economic Research Service The Economic Research Service, working with NASS, annually col- lects data on farm costs and returns, land values, and resource and en- vironmental aspects of farm production practices such as fertilizer and pesticide use. Surveys are designed jointly with NASS, and NASS enu- merators collect the data in regular and special surveys. Geographic in- formation systems and farm record databases developed on the farm could provide information superior to current surveys because they would provide data on soils, weather, and other important variables integrated directly with the economic data. Currently, physical factors affecting eco- nomic decision making must be inferred from other data sources. National Resources Conservation Service The Natural Resources Conservation Service (in addition to the Co- operative Soil Survey) conducts the National Resources Inventory (NRI), an area-based statistical sample of land cover and use, soil erosion, continued on next page

116 PRECISION AGRICULTURE IN THE 21ST CENTURY BOX 4-1 Continued prime farmland, wetlands, and other natural resource characteristics on nonfederal rural land in the United States (excluding Alaska). Inventories such as the NRI have been conducted since 1945 and are now con- ducted at five-year intervals. The 1992 NRI is the most extensive inventory yet conducted, covering some 800,000 sample sites and representing some 75 percent of the nation’s land area. Many of the data elements and definitions used to collect the 1992 data were developed to be comparable with data con- tained in the Commerce Department’s Census of Agriculture and with databases managed by the USDA Forest Service, USDA National Agri- cultural Statistics Service, and the Interior Department’s U.S. Geological Survey and U.S. Fish and Wildlife Service. Precision GIS and farm record databases could provide information superior to that of the current National Resources Inventory because they could provide data on actual soil qualities such as nutrient content, or- ganic matter, pH, and electrical conductivity directly at the site, rather than through inference from typical properties of that soil type. Precision farm data could also integrate economic data on input use, yields, and production with the physical data. Currently, data on inputs and outputs must be inferred from other data sources. not have enough detail for site-specific decision support (i.e., precision agricul- ture, construction suitability, septic disposal, and land filling). On the other hand, these spatial scales have too much detail for more summary analysis (i.e., land use planning, land suitability, and regional groundwater analyses). Generation and automation of data at this intermediate scale may be wasted effort that satis- fies few potential clients. NCSS does not address precision agriculture’s requirements for soils data. Soil interpretations provide typical characteristics of soils at surveyed sites but do not record observations of the characteristics of individual soils at these loca- tions. The orientation is toward the other end of the detail-resolution spectrum, providing data for resource and environmental applications across extensive ar- eas. In their examination of aspects of NRCS data activities, a blue ribbon panel articulated similar recommendations for changes within NRCS (Natural Re- sources Conservation Service, 1995). Under the leadership of NRCS, NCSS has been making some steps in the right direction. For example, the Soil Survey Program Plan includes surveys that document information on soil landscape relations as well as soil taxonomy. This

PUBLIC POLICY AND PRECISION AGRICULTURE 117 should result in better information on the reliability and variance structure of soil data. NRCS has been developing the National Soil Information System, a set of GIS and statistical tools for providing access, analysis, and manipulation of county-based soil survey information, including digital spatial data. The shift in orientation toward characterizing soil landscape relations and soil variability should benefit some kinds of precision agriculture (i.e., systems that include ma- nipulation of soil-water relations through modeling of surface hydrology and soil characteristics). It is likely to be many years, however, before this kind of infor- mation is widely available and useable in a management context. For NRCS and NCSS to provide useful information for a broad spectrum of precision agriculture, they will need to carefully review the needs of precision agriculture systems and methods. It is probably not appropriate or even feasible for NCSS to map soils at the level of detail required for many types of precision agriculture or to collect detailed soil characteristics in a fine grid. This activity has little public benefit and there is little justification for public investment. How- ever, NCSS could play an important role in providing the information infrastruc- ture for such detailed work by developing (a) data quality standards for detailed work in addition to the standards for development and automation of their current products; (b)methods for data collection, testing, and interpretation; and (c) pro- cedures for accessing and archiving data by private soil consultants. NRCS and other public agencies could also benefit directly from more de- tailed data collected on farms. Farm-generated data could be used to more effec- tively characterize soil variability and soil landscape relations within a region such as a major land resource area; in effect, the farm would serve as a research site. Such a public-private partnership would require NCSS to interact with a new group of users and to relax their push for uniform products. This scenario is based on the assumption that privacy and intellectual property rights issues could be resolved in such a way as to allow NRCS and NCSS access to at least portions of data collected privately. Site-specific data might be used only in model develop- ment and statistical inference and thus be made generally available only in aggre- gate or processed forms (this may require a Freedom of Information Act exclu- sion for the microdata about specific farms held by NCSS). Issues parallel to those for collection of soil survey data exist with other data collected on farms by federal agencies. Agricultural statistical agencies should take steps now to assess the likelihood and speed of development of precision agriculture data collection and devise approaches to tap the enormous potential of such data flows. Pilot projects to develop data warehousing techniques and proto- cols could yield large dividends in accommodating an eventual shift from survey- to precision-based data collection. As precision agriculture becomes more widely adopted, precision data could at first supplement, and perhaps later entirely sup- plant, more traditional data collection paradigms based on agency surveys. Federal agencies cannot immediately use precision agricultural data because the number of operators who have fully adopted precision agriculture and thus

118 PRECISION AGRICULTURE IN THE 21ST CENTURY the acreage covered are minimal. Even if precision agriculture were widely adopted, numerous obstacles must be overcome before the data collection para- digm could change as described above. For agencies to use producer-collected data, some institution will have to impose metadata standards specifying a minimum level of consistency in content, format, and protocols which producer-collected data must meet. This is not a trivial problem, as evidenced by the amount of time and coordinating efforts re- quired to develop a spatial metadata standard for the data that federal agencies already collect themselves (Federal Geographic Data Committee, 1994; Federal Register, 1994; National Research Council, 1994). Whether government agen- cies take the initiative to develop data warehousing systems or cooperative or private enterprises emerge, order must be imposed to prevent the chaos that could result from simply gathering what individual producers collect. Developing standards is fraught with many problems. First, the technologies used in precision farming are evolving so rapidly that standards may always lag implementation in the field. Second, companies in the hardware and software industries providing precision agriculture technology may want to maintain pro- prietary standards wherever there is a competitive advantage to do so. Encourag- ing developments from the Agriculture Electronics Association show, however, that there can be industry-wide cooperation on standards. Finally, the problems of data ownership, data privacy, and data sharing discussed above may limit produc- ers’ willingness to contribute data or to be bound by any standard. If such a metadata standard does evolve, several legal barriers could thwart use of individual producers’ databases. Protecting data confidentiality is a matter all federal statistical agencies take seriously. Access to microdata records and safeguards against disclosing an individual’s identity in summary statistics are already issues with agencies collecting data. The problems will be much more complex if producers provide data directly from their own computer records. Any systematic use of precision agriculture data must safeguard the producer not only from the general public and competitors, but from other federal and state agencies that exercise regulatory and taxing power. Producers’ precision data will not be forthcoming if the data can be used by the EPA to fine polluters or the Internal Revenue Service to second-guess tax returns. Property rights concerning the data are also a consideration and a potential barrier. To the extent that producers perceive that there is economic value to the data beyond their farm gate, they could require payment for use of the data. Even if producers are willing to contribute data to the common good, they incur costs for collecting the data, developing systems to record and store the data, and trans- mitting the data to warehouses; the warehouse may then require reimbursement by agencies that use the data. Accommodating this reality will require a large adjustment on the part of agency administrators who currently obtain survey data from producers at no direct cost. Finally, the sheer size of databases that could be developed from precision

PUBLIC POLICY AND PRECISION AGRICULTURE 119 farm data collection is a barrier to agency use. The process of turning the result- ant mountains of raw data into useable and useful information, without sacrific- ing its inherent geographic specificity and detail, is a formidable challenge unlike the one currently facing agricultural data agencies. The limiting constraint today is additional resources for more samples, whereas the limiting constraint in a precision agriculture data world may be the computer methods and power needed to store, process, and summarize the available data. The vaults of LANDSAT data tapes residing at the USGS Earth Resources Observation System data center in Sioux Falls, South Dakota, provide an instructive analogy. The raw download of data from almost 20 years of satellite operations, even though accessible, is so daunting a processing and interpretation task that only a small fraction of these data have been converted into relevant information. Summarizing the flood of data that could derive from two million precision farm databases would be that much more daunting. POTENTIAL FOR PRECISION AGRICULTURE The committee believes that precision agriculture offers new information technologies to address information needs for management of agricultural crops. Widespread adoption of precision agriculture technologies will constitute a new way to practice agriculture at ever finer spatial and temporal resolutions, and to improve use of information for crop management at all spatial scales. These new capabilities offer the potential for a more economically and environmentally effi- cient agricultural sector. However, precision agriculture technology is new and largely unproven. Widespread adoption depends on economic gains outstripping the costs of the technology. Exploiting the full potential of precision agriculture for environmental management will require fundamental shifts in public and pri- vate incentives for environmental management, and may require cost-sharing or other incentives for adoption. Lessons from the adoption of other agricultural and information technologies urge caution in anticipating the growth of precision ag- riculture use. Widespread adoption of precision agriculture methods will create changes in farm operations and in social institutions that can be anticipated and, where they are negative, mitigated. Many of the important findings in this report deal with the range of public policy responses to precision agriculture’s evolution and adoption.

Next: References »
Precision Agriculture in the 21st Century: Geospatial and Information Technologies in Crop Management Get This Book
×
Buy Paperback | $52.95 Buy Hardback | $52.95
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Sensors, satellite photography, and multispectral imaging are associated with futuristic space and communications science. Increasingly, however, they are considered part of the future of agriculture. The use of advanced technologies for crop production is known as precision agriculture, and its rapid emergence means the potential for revolutionary change throughout the agricultural sector.

Precision Agriculture in the 21st Century provides an overview of the specific technologies and practices under the umbrella of precision agriculture, exploring the full implications of their adoption by farmers and agricultural managers. The volume discusses how precision agriculture could dramatically affect decisionmaking in irrigation, crop selection, pest management, environmental issues, and pricing and market conditions. It also examines the geographical dimensions—farm, regional, national—of precision agriculture and looks at how quickly and how widely the agricultural community can be expected to adopt the new information technologies.

Precision Agriculture in the 21st Century highlights both the uncertainties and the exciting possibilities of this emerging approach to farming. This book will be important to anyone concerned about the future of agriculture: policymakers, regulators, scientists, farmers, educators, students, and suppliers to the agricultural industry.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!