Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 393
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS Appendix A Poster Sessions Low-Input Sustainable Agriculture Farm Decision Support System John E. Ikerd U.S. farmers are faced with growing environmental concerns and rising costs associated with highly specialized farming operations. They are searching for farming systems that are ecologically sustainable as well as productive and profitable. Many are motivated by perceived risks that the inputs on which they depend today may not be available, may not be effective, or may cost much more in the future. Such farmers are searching for ways to reduce their dependence on external purchased inputs while maintaining their productivity and profits through more intensive management of their internal resources. The current search for sustainability and profitability in U.S. agriculture is centered on helping farmers develop more ecologically sound and economically viable farming systems with existing technology while searching for even more sustainable and profitable alternatives for the future. A short-term objective is to improve the input efficiency of current farming systems. However, long-term sustainability may require more diversified systems of farming that include commodities that can be produced with more ecologically benign systems. Diversified farming systems traditionally use crop rotations to control pests, conserve soil, and maintain productivity. Integrated cropping and livestock systems have been used to reduce feed costs, recycle waste, and stabilize the incomes of U.S. farmers. The current hope for future success lies in finding ways of combining
OCR for page 394
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS new technologies such as microcomputers and biotechnology with the tried and proven principles of management by objectives and diversification —old principles with new technologies. Such farming systems will be more complex and thus will require more intensive hands-on resource management than do higher-input, specialized systems. However, synergistic gains from effective integration of enterprises and activities in diversified farming systems represent the best hope for achieving long-term sustainability with a minimum of government regulation. A microcomputer-based farm decision support system1 is being developed under a project funded jointly by the Extension Service and Cooperative State Research Service, U.S. Department of Agriculture, to integrate the concept of sustainability into farm planning and to implement farm management strategies for sustainability. The Low-Input Sustainable Agriculture Farm Decision Support System (LISA-FDSS) project was approved in 1988 for funding through November 1990. CHARACTERISTICS OF LISA-FDSS2 The LISA-FDSS system has six basic functions that are supported by two microcomputer-based program components, two farming systems data bases, and several specialized data bases to support the budgeting process. LISA-FDSS is designed to be compatible with a national financial planning project, FINPACK, and a national linear programming project that emphasizes labor and machinery management. The six basic functions of the LISA-FDSS system are as follows: (1) resource management strategy (RMS) budgeting, (2) whole-farm planning, (3) environmental checking, (4) financial checking, (5) risk checking, and (6) resource checking. The two data bases are (1) default RMS budgets and (2) customized RMS budgets. Additional data bases include soil types and characteristics, fertilizer and pesticide characteristics, correlation coefficients, and energy conversion units. RMS Budgeting The RMS associated with a cropping system consists of a crop sequence or rotation, an irrigation system (if any), a tillage system, a fertility system, and a pest management system. An RMS budget reflects the resource requirements, input requirements, input costs, expected production, expected returns, potential conservation impacts, and potential environmental impacts of the individual crops as components of a cropping system. An RMS budget contains all non-site-specific information needed to calculate
OCR for page 395
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS expected soil losses, water-quality risks, resource use, gross margin over purchased inputs, and revenue risks. The default RMS budget data base will contain budgets for cropping and livestock systems deemed appropriate for the geographic region of application. These data bases will be constructed by extension specialists using a basic FINPACK financial budget format augmented by additional resource and environmental (R&E) components. Development of R&E budget components will be facilitated by a budgeting program developed as a part of the LISA-FDSS project. The R&E budgeting program is one of the two basic microcomputer program components of the LISA-FDSS system. Default data bases should include budgets for a wide range of cropping systems deemed appropriate for the geographic region where the LISA-FDSS program is to be used. A cropping system might include from 1 to 12 different crops. A monocrop system would have the same budget for each year. A given crop following different crops in different rotations might have a different budget for each rotational position. Different crops, of course, would have different budgets. Each cropping system will be budgeted for up to four alternative input systems. An input system will reflect a specific fertility and pest management system. Most systems would be budgeted with unrestricted-input, reduced-input, and low-input RMS alternatives. Unrestricted-input RMS budgets will reflect the use of typical fertilizer and pesticide inputs for a particular cropping system on fields with no significant fertilizer or pesticide leaching or runoff risk potential. Reduced-input RMS budgets will reflect some lower level of inputs suggested for fields with significant nutrient or pesticide risk potentials. Split applications and banding of fertilizers and pesticides might be a logical reduced-input system, for example. A low-input system should reflect minimum levels of external inputs that specialists deem feasible for commercial production on fields with high nutrient loading or pesticide risks. Each cropping system will also be budgeted for alternative tillage levels. Tillage options will range from unrestricted tillage to minimum tillage. Unrestricted tillage would be the suggested system for fields without erosion problems, with minimum tillage suggested for highly erodible fields. Each tillage system should be matched with an appropriate complement of inputs. Consequently, some systems may have no low-input, minimum-tillage RMS, if such a combination of tillage and inputs is not considered feasible for a given cropping system. In general, the alternative input systems will be designed to reduce water-quality and other environmental risks by moving to lower-input alternatives. In general, the alternative tillage systems will be designed to reduce soil erosion risks by moving to lower tillage levels. Irrigation systems, if any, will be specified as a part of each input system.
OCR for page 396
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS Farmers who use an unrestricted system for one crop would likely use an unrestricted system for another crop in the same rotation, although an unrestricted system might imply different tillage and input regiments for different crops in the same rotation. Likewise, a farmer interested in a low-input commercial alternative for one crop likely would be interested in a similar system for other crops in the same rotation. Thus, the levels of inputs and tillage will be identified for whole cropping systems rather than individual crops. Whole-Farm Planning The Whole-Farm Planner (WFP) is a microcomputer-based decision support system that allows farmers to evaluate the potential impact of using various cropping systems or RMSs on their specific farms. The WFP is a field-based system. It allows farmers to plan their farms field by field and year by year and to assess the RMS implications for each field and each year for the whole farming system, including livestock as well as crops. A typical FDSS user would begin with the whole-farm planner component of the system. An agent working with a farmer should have determined the basic rotations used by the farmer and have those RMSs available in the default data base at the time of the first planning session. Otherwise, the farmer and agent would have to add those budgets to the default data base before the planning process could begin. Most farmers will want to begin with an assessment of their current system before they begin to evaluate alternatives. All site-specific information and the associated yield and environmental impact estimates are calculated within the whole-farm planner program. Thus, the whole-farm planning process begins with a field-by field inventory of the land or soil resources of the farm. Much of the information related to soil erosion and environmental vulnerability can be derived from the Soil Conservation Service (SCS) data base of soil types. Soil texture, pesticide leachability, pesticide surface loss potential, and average slope and slope length are identified in the SCS data base of U.S. soils. However, the farmer will be asked to verify yield potentials, soil characteristics, and environmental impact estimates in the planning process. Environmental and conservation impacts will be evaluated field by field over a 12-year planning period. Thus, estimates of soil loss, water-quality risks from pesticides and fertilizers, and input toxicity will be evaluated for cropping systems rather than individual crops. Financial and resource implications of alternative systems will be evaluated for the whole farm for each year in the planning period. The acreage of each crop, pasture, set-aside or conservation reserves, expected revenues, input costs, gross margins, revenue risks, corn equivalents produced and
OCR for page 397
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS needed, hay equivalents produced and needed, and nonrenewable energy use will be summarized for each year. The ecological vulnerability of each field will be identified by color-highlighted codes for high, medium, and low levels of vulnerability to soil loss, pesticide leaching, and residue runoff. Each cropping system and RMS will likewise be color-coded with respect to its potential for soil loss and water-quality risks. These two sets of codes, one for the field and the other for the RMS, will be combined to yield a similar color-coded set of implications for using a given RMS on a given field. Each combination of field and RMS will have a color-coded indicator of soil loss, water-quality risk from pesticide and nitrogen use, and input toxicity. A set of red H's for a given RMS on a given field, for example, could indicate severe ecological problems. Such problems would be associated with the use of a particular RMS on a particular field. The same RMS might be acceptable on another field, but a different RMS might be indicated for the particular field being examined. There will be relatively few alternatives for correcting the ecological vulnerability of a given field. Exceptions would be to contour till, terrace, strip crop, or ridge till a field to reduce soil loss potential. In most cases, farmers will have to change RMSs to correct ecological problems. Each RMS will be identified with a code indicating the tillage and input levels associated with the particular strategy. A farmer with an erosion problem might consider an RMS with less tillage. If, instead, the farmer is faced with a water-quality problem, he or she might select a lower-input RMS. If the farmer has an erosion and water-quality problem, he or she could select a longer crop rotation that included meadow or some other soil-conserving crop. A similar approach will be used in the financial, risk, and resource sections of the program. An unacceptable income level for a given year would be color coded with a red H or some similar sign. The farmer could first consider shifting rotations to get more high-income crops in a given year, if the problem occurred only for 1 or 2 years. However, if the problem occurs for several years, he or she may consider some more intensive RMSs that will generate more income in more years. Inconsistencies between labor needs and availability would be flagged. Seasonal labor problems may be addressed by shifting rotations, changing to lower-labor RMSs, or hiring labor during peak need periods, if it is feasible. Feed needs and production would be handled in a similar manner. An unacceptable level of risk might suggest that diversity be added by selecting alternative cropping systems, adding livestock to the system, or, possibly, considering off-farm employment for income stability. Changes in RMSs to solve financial, risk, or resource problems may generate ecological problems. However, no attempt will be made to calcu-
OCR for page 398
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS late an optimum system for a given farm. Farmers will simply attempt to solve their ecological and economic problems by matching alternative resource management strategies (cropping systems with alternative tillage and input levels and livestock enterprises) with their internal resources (land, labor, and machinery). Information describing each RMS, including any specialized machinery requirements, will be available from the whole-farm planner program. For example, a farmer may want to know what type of fertility program, tillage system, pest control system, and labor requirements are assumed for a low-input soybean alternative in a corn-soybean rotation in field number three in year 4 of the plan currently on the screen. He or she would indicate with some set of key strokes the basic data he or she wants to review for this particular alternative. The whole-farm planner program assumes that a farmer has multiple objectives that include both ecologic and economic factors. The ecologic factors are soil loss, water quality, input toxicity, and nonrenewable energy use. Standards for the ecological factors will be predetermined. The economic factors are net returns or income; income risks; and utilization of land, labor, and machinery. Farmers will be asked to develop their own income objectives from overall farm financial information. Some farmers may be willing to settle for a whole-farm plan with a large number of red, or warning, indicators on the ecologic factors to achieve green, or safe, indicators in the financial and resource areas. Others may be willing to tolerate lower economic results to achieve safe indicators (green color codes) in the ecologic areas. Others will continue to explore alternatives until they have all ecologic and economic indicators in acceptable ranges or they will not farm. These choices are to made by the individual farmer. Custom Budgets Each farmer would need to work with his or her agent or specialist in customizing the default RMS budgets to reflect inputs and resources for tillage and cropping systems that the farmer actually expects to use on his or her farm. The WFP program would allow the farmer to greatly narrow the range of budgets that might be considered to be logical for his or her operation. However, the customized alternatives need not be limited to those for a single best farm plan identified by the farm planner. Changes from default values to customized values for environmental and economic impacts may significantly change the estimated outcomes of a given farm plan. Thus, once the customization process is completed, the farmer would be expected to return to the WFP program. He or she would simply repeat the earlier iterative planning process with the customized sets of budgets until a satisfactory customized plan is achieved.
OCR for page 399
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS OBJECTIVES OF LISA-FDSS The RMS budgeting process will allow agricultural specialists to reflect the full range of existing and future research results and information in a form that is readily usable by farmers. For example, ecologic and economic impacts of cover crops, intercropping, and relay cropping in various rotations can be reflected in alternative RMS budgets. Uses of legumes and livestock manure for fertilizers as well as alternative systems of fertilizer application can be included among the RMS alternatives to be considered. Impacts of alternative tillage systems and residue management programs on potential soil loss will be an integral part of the budgeting process. Alternative weed, insect, and other pest control systems, including specific pesticide uses and their potential risks to humans and water quality, will be reflected directly in the environmental components of each RMS budget. The whole-farm planning process will allow farmers to synthesize profitable and sustainable farming systems by integrating relevant RMSs with their particular set of land, labor, machinery, and management resources. They can select RMSs that are well-suited for their soils, climate, and location-specific pest problems. They can integrate systems of livestock and crop RMSs that tighten or complete nutrient cycles, facilitate energy flows, and enhance the ecologic and economic viability of their farming systems. Farmers who use the whole-farm planner can evaluate potential impacts of using various levels of various chemical fertilizers and pesticides on specific fields. They can match tillage systems and soil-conserving practices with specific slope and soil characteristics of fields to reduce erosion. They can assess risks through evaluation of diversification effects of alternative farming systems and develop systems that are resistant, resilient, and regenerative. The LISA-FDSS will not result in a recipe for success. LISA-FDSS is just a tool to facilitate farm planning and management. Farmers who choose an alternative to their current system will be advised to gather as much additional information as is available before they adopt a new farming enterprise or practice. Farmers will be strongly encouraged to talk with other farmers who have experience with the practice under consideration. They will be encouraged to visit other farms where the practice is used before they change their own operation. They will be advised to work into any new system slowly, so they can learn as they go. The LISA-FDSS will not ensure a more profitable or sustainable farming system. However, it will allow farmers to evaluate the potential impact of alternative LISA technologies and strategies within the context of their
OCR for page 400
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS particular farming situation without doing the necessary research and testing on their own. Use of the LISA-FDSS will not ensure success, but the LISA-FDSS can be a valuable and important aid in taking the first step toward the goals of economic and ecologic sustainability. NOTES 1. This farm decision support system was developed by the LISA-FDSS Task Force: John Ikerd, Columbia, Missouri; Richard Levins, St. Paul, Minnesota; Larry Bond, Logan, Utah; Mike Duffy, Ames, Iowa; Don Tilmon, Newark, Delaware; Tim Hewitt, Marianna, Florida; and Patrick Madden, Glendale, California; special funding was provided by the Extension Service of USDA. 2. LISA-FDSS has been renamed Sustaining and Managing Agricultural Resources for the Future—Farm Resource Management System (SMART-FRMS). Further development and support of the system is being carried out by the Center for Farm Financial Management, University of Minnesota, St. Paul, Minnesota. Voisin Controlled Grazing Management: A Better Way to Farm William M. Murphy Permanent pastures in the northeastern United States typically have low productivity, producing only about 2 tons of moderate-to-poor-quality forage per acre during a 3- to 4-month grazing season. A proven method exists that enables these kinds of pastures to produce 4 tons or more of excellent-quality (23 percent crude protein, 0.72 Mcal/pound of net energy lactation) dry forage per acre during a 6- to 7-month grazing season. The method is controlled grazing management, as described by Andre Voisin (1959) (see also Murphy, 1987). This method, which is also known as short-duration grazing, intensive rotational grazing, and rational grazing, has been used for many years in New Zealand and for 8 years in Vermont. New Zealand's highly productive and profitable agriculture depends almost entirely on permanent pastures that are grazed under controlled management. New Zealanders raise 70 million sheep, 8 million cattle, 1 million deer, and 1 million goats, without grain supplements, on only 37 million acres of pastureland, which is the size of Iowa. This proves that the method works. Many American dairy farmers, in contrast, use a system of zero pasturing or year-round confinement feeding that involves large amounts of purchased feed and supplements, huge capital investments in facilities and
OCR for page 401
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS equipment, large cash flows, and low profitability. Partly because of this, many U.S. dairy farmers are experiencing a financial crisis that may eliminate many family farms, because feeding of livestock in confinement can cost six times as much as it does on well-managed pastures. American farmers who do not use year-round confinement feeding put their livestock to pasture, where the animals are grazed continuously or rotated through a few large pasture divisions with little control and less planning. Invariably, by late June or early July the pastures are depleted and worn out. These dairy farmers generally do not feel that their pastures have much feed value and use the same ration all year, regardless of what the pastures produce. Therefore, pastures have been a wasted resource in the United States. One way to increase profitability of a farm is to reduce feed costs. The permanent pastures that exist on most farms produce biomass far below their potential because of poor grazing management. Pastures managed under controlled grazing conditions can be some of the most valuable areas on a farm, producing high yields of excellent-quality forage. When incorporated into livestock feeding programs, this homegrown forage can reduce feed costs and increase the profitability of many northeastern farms. First-year costs of materials, maintenance, and labor for the grazing management method range from $1,500 to $2,000 for a 40-cow herd. Its use has returned $3.75 in benefits for each $1 invested by dairy farmers (Jones and Burns, 1988; Pillsbury and Burns, 1989). Voisin grazing management is a simple system of controlling grazing by dividing pastures into small areas (paddocks) that are grazed on a rotational basis. This method minimizes the waste of forage and protects the plants from overgrazing. GUIDELINES FOR VOISIN CONTROLLED GRAZING MANAGEMENT IN VERMONT The essentials of the Voisin grazing method and what its use has meant to three farmers in terms of increased profitability and improved quality of life are illustrated in a 33-minute video produced as part of a low-input sustainable agriculture (LISA) project (Murphy et al., 1989). Recovery Periods The recovery periods between grazings must vary with the plant growth rate. This usually means that recovery periods must increase as the plant growth rate decreases as the season progresses. In Vermont, for example, this means that a 10- to 18-day recovery period is needed during May and June. This gradually increases to 36- to 42-day recovery periods by
OCR for page 402
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS the end of September. Another way to look at it would be as follows: 10 to 12 days of recovery time in late April to early May, 15 to 18 days by June 1, 24 days by July 1, and 36 days by September 1. These are guidelines only; longer or shorter recovery times may be needed, depending on the local growing conditions. Recovery periods reflect the pre- and postgrazing pasture mass (total forage) relationships shown in Figure A-1. Pasture mass influences most the net harvested forage production at the extremes of low postgrazing and high pregrazing masses. At a low pasture mass, the lack of leaf surface area limits solar interception and photosynthesis. At a high pasture mass, shading of lower leaf surfaces blocks solar interception, while respiration of shaded plant parts consumes the carbohydrates that are produced, until death and decomposition of the shaded parts occur, with consequent loss to net production. Based on these relationships, the forage should not be taller than 6 to 8 inches when cows are turned into a paddock (the forage should not be taller than 4 inches for sheep, because sheep-grazed swards are more dense) FIGURE A-1 Effect of pasture mass (as dry matter [DM]) on rates of new herbage formation, net forage production, and forage losses. Forage losses through death and decay result mainly from shading of lower plant parts, and these losses increase as pasture mass increases. Source: C. J. Korte, A. C. P. Chu, and T. R. O. Field. 1987. Pasture production. Pp. 7–20 in Feeding Livestock on Pasture, A. M. Nichol, ed. Occasional Publication No. 10. Hamilton, New Zealand: New Zealand Society of Animal Production.
OCR for page 403
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS and should be grazed down to 1 to 1.5 inches from the soil surface before the animals are removed. If the animals do not eat enough to keep up with the rapidly growing forage in the spring, some of the paddocks should be removed from the rotation and should be harvested for hay or silage. Usually, about one-half of the pasture area must be saved for machine harvesting, because too much forage is produced in May and June. This means, for example, that if there are 20 paddocks, 10 of the most level ones should be saved for machine harvesting. After harvesting, the paddocks should be rested until the plants regrow adequately before they are included in the next rotation. The larger number of paddocks then available for grazing in late July and early August automatically increases the recovery periods of all paddocks. If, at any time, the paddocks have not fully recovered by their turn in the rotation, all of the animals should be removed from the pasture and should temporarily be fed elsewhere (e.g., they could be grazed on the hayland aftermath or fed green chop, hay, or silage harvested from the excess earlier in the season) until recovery periods are adequate before the animals are turned back into the pasture system. By strictly observing this need for adequate recovery times, permanent pastures in areas such as Vermont may be able to be grazed from mid-April to mid-November. In contrast, pastures that are not under Voisin controlled management can be grazed for a much shorter time, typically from mid-May to mid-August. Periods of Occupation The total time that animals occupy a paddock in any one rotation must be less than 6 days, to prevent grazing of regrowth in the same rotation. Paddocks must be small enough so that all or most forage in each paddock is grazed down to about 1.5 inches from the soil surface within this time limit. If two separate groups of animals are grazed (e.g., milking cows or heifers, dry cows, and lambs and ewes), each group should not be in a paddock for longer than 3 days, because forage palatability and availability decrease too much after 3 days for each group. In practice, the shorter are the periods of occupation, the better it is for optimum plant and livestock production. If animals are grazed as one group, they should not be in a paddock for longer than 2 days for the best livestock production. If two groups graze a paddock, each one should be in the paddock for only 1 day. Milking or fattening animals should not be in a paddock for longer than 1 day, so that they can be kept on a consistently high level of nutrition. Milking cows produce the most if they are given a fresh paddock after every milking. Growing lambs and beef cattle should be moved to a fresh paddock once a day for the best results.
OCR for page 417
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS troubleshooting and may be doing more of the double checking of their own knowledge mentioned above. These measures of system use taken together indicate one aspect of adoption: use of the innovation. While the number of times that the PSAOC system is accessed shows how frequently the system is being used, the actual amount of time spent using the system may be a more significant indicator of adoption of the innovation. Some growers reported that they used the system primarily as a quick validation of their own knowledge regarding a decision. These growers reported a relatively high number of accesses and a low number of hours used. Conversely, the growers who reported that they used the system for many hours were presumably more fully engaging the logic of the system in their decision-making process. General Practice Change Characteristics The degree to which growers follow the recommendations presented by the expert system is a second aspect of adoption. Table B-2 displays two measures of the frequencies of changes induced by use of the system: (1) any change in growers' production practices and (2) increased pest monitoring. Both measures were derived from the eight monthly surveys. The first measure is a sum of the number of times that growers indicated that use of the expert system stimulated some change in their production practices. Over the course of the 8 survey months, 65.2 percent of the growers indicated that they had changed standard production practices in some way during at least 1 month. Of these growers, 17.4 percent indicated some change during 3 different months of the 8 survey months. A significant number of those sampled (65.2 percent) engaged a new and untried technology and were stimulated to change production practices as a result. The second of the practice change characteristics displayed in Table B-2 is a sum of the number of times that a grower was stimulated by the expert system to go to the orchard and scout for a pest (monitoring). Pest monitoring is seminal to any IPM program. A large majority of growers (82.6 percent) reported that the system stimulated them to increase their monitoring at least once. A total of 30.3 percent of growers were stimulated to monitor their orchards four or more times. As the majority of pest monitoring occurs during April, May, and June, these numbers take on more significance when viewed as a subset of the eight monthly observations. Weekly Time Monitoring and Basic Economic Questionnaires During the field test and evaluation process in the 1989 season, the economic impact of the apple expert system on cooperators' operations and net
OCR for page 418
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS TABLE B-2 Penn State Apple Orchard Consultant Expert System Adoption Characteristics of Growers Production Practice Change Characteristic Percentage of Growers (n = 23) No. of times growers reported some change in practices, per grower 0 34.8 1 21.7 2 26.1 3 17.4 No. of times system stimulated increased pest monitoring, per grower 0 17.4 1 26.0 2 4.4 3 21.7 4 21.7 6 4.4 7 4.4 income was estimated. Many growers already maintain pesticide logs that contain most of the data needed for development of an apple enterprise budget. A basic economic survey questionnaire was developed from the pesticide record and crop history logsheet of a major commercial apple processor to collect orchard characteristics, apple yields, and prices received. Additional information to aid in the comparison between expert systems users and a control group of nonusers was incorporated into the questionnaire. A weekly time monitoring survey was designed to gather information on the amount of time each grower spent scouting (monitoring) his or her orchard each week as well as what pest problem was being looked for. Pesticide application records were also collected to provide information on the chemicals and rates that the chemicals were applied to each orchard. The survey questionnaire was subjected to three reviews: first, by the research team; next, by all the county agents involved in the project; and finally, by selected growers who had expressed interest in its development. This feedback was particularly helpful for developing the yield and price components of the questionnaire, which was a two-part format that was collected in the spring and the fall. Results from the monitoring surveys are still being analyzed. While the findings reported here are preliminary and subject to change, they, too, indicate that the expert system is an effective teaching tool. In the past,
OCR for page 419
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS extension information has encouraged growers to monitor for mites at the time of bloom and thereafter (week 8 of the growing season). Both PSAOC users and nonusers performed scouting at similar frequencies in the postbloom period. However, a new prebloom monitoring practice is recommended by the expert system as an effective mite control strategy that may reduce pesticide usage later in the season. The nonusers of PSAOC were not as aware of this prebloom method. Figure B-3 shows that more PSAOC system users tended to monitor for European red mites earlier in the season than did the comparison group of nonusers. Similar behavior has been seen in PSAOC users who ended their monitoring processes sooner than did the control group, thus making more efficient use of limited time. This constitutes direct evidence that use of an expert system can stimulate measurable changes in farming practices. A preliminary comparison of the farm-gate economics of expert system users versus those of expert system nonusers shows some trends. Even though Pennsylvania suffered through a poor apple-growing season in 1989, the preliminary results of the survey show that yields of PSAOC users and nonusers were roughly similar. The cost of time spent monitoring the orchard for pests and using the expert system is also a component of the economic impact being examined. Specifically, the team is looking to answer the question of whether savings FIGURE B-3 Monitoring for European red mites (ERM) by users and nonusers of the Penn State Apple Orchard Consultant expert system.
OCR for page 420
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS on pesticide applications were being offset by greater costs in management. A weekly time-monitoring survey was developed that provides a checklist for most of the common items monitored. Primarily, it asks how much time was spent monitoring each block and using the expert system. This checklist went through the same review process the basic economics survey did. No clear results have yet been obtained from the pesticide records analyzed thus far, but some interesting trends have been noted. There is some indication that system users may have applied lower amounts of some insecticides than nonusers did. Further analysis of this information may indicate whether or not the expert system is changing growers' practices regarding pesticide use and will provide the basis for partial budget analysis. Further Mechanisms to Obtain Grower Evaluation, Feedback, and Training Cooperators' Planning and Review Meetings The experiences with the PSAOC expert system during the 1988 and 1989 growing seasons were summarized during facilitated meetings of cooperating growers, researchers, and extension personnel in February 1989 and March 1990, respectively. The primary purposes of the meetings were to review the system's performance over the year to date, provide the growers with an opportunity for in-depth input and discussion about improvements in the program, and collectively plan for the upcoming year. In addition, a major benefit was to bring growers from 13 counties in Pennsylvania and researchers and extension agents from three states together to interact for the first time. The nominal group technique was employed during working sessions with the growers group to solicit any suggestions that they had for improving either the software itself or the field evaluation process. Recommendations were distilled and ranked by growers according to importance during a later session. Growers and extension agents also strongly suggested the inclusion of more economic information into the PSAOC expert system. A session devoted to procedures for collecting relevant budget data yielded an additional step in the proposed analysis of farm-level economic impacts. Researchers and extension specialists from The Pennsylvania State University (University Park), University of Massachusetts (Amherst), University of Vermont (Burlington), and the Rodale Research Center (Maxatawney, Pennslyvania) also met for 1.5 days to plan and coordinate the following year's program. Additional responsibilities for expert systems
OCR for page 421
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS development and evaluation were outlined for the second- and third-year plans of work. Midseason Grower Training Sessions Based on feedback from growers as well as trends in the survey data, small-group training sessions were held at the Biglerville Fruit Laboratory and the Berks County Agriculture Center during the summer of 1989. It was determined that the newest version of PSAOC was not being comprehended adequately and therefore was not being used to its fullest efficiency. These training sessions sought to correct this problem by familiarizing the growers in-depth with the new aspects of the software. Electronic Mail Network Among Growers and Researchers Also in response to feedback from growers, an electronic mail users group was formed to improve communications between cooperating growers, researchers, and extension personnel. Using The Pennsylvania State University's PenMail system, the growers are able to communicate with each other, with county extension agents, and with specialists on campus via electronic mail. This communications link has helped to make growers more comfortable with the computer and the information they receive. The electronic mail system was set up in March 1989. Grower communications have included questions about insects, pest trapping, use of the computer, and information on the new version of PSAOC. The project 's evaluation coordinator has sent out numerous informational and update bulletins. The growers are also receiving their own copy of the state horticultural newsletter by electronic mail. Half of the growers have accessed the system (for messages, responses, PenMail) roughly once a week, and the others have accessed the system about once a month. This system has worked well so far, and it is expected that usage will continue to grow. Site Visits to Cooperating Orchards Visits to field test sites were made by evaluation staff at various points during the growing season, to observe orchard management and expert systems use by grower. These visits also provided more opportunity for the growers to give input into the development and improvement of PSAOC. It was noted that the expert system was more often found in the business office of the orchard, residing on the computer the grower used for accounting. Grower Panel at Professional Meetings Three pilot study growers and the cooperating regional tree fruit exten-
OCR for page 422
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS sion agent presented a panel discussion on the Penn State Apple Orchard Consultant system to over 300 apple growers at a meeting of the Pennsylvania Horticultural Association in January 1990. Discussants provided insights into their experiences with testing of the system, citing both the problems and potentials of using the expert system in orchard management. Panelists were mostly supportive of the new technology, citing increased responsibility on the part of the grower to reduce environmental inputs and improve food safety while still maintaining profitability. Involvement with Cooperative Extension Agents Cooperative extension agents were directly involved in the organization and implementation of the project. In addition to consulting on the structure and content of the survey process, agents were primarily responsible for the selection of cooperating growers for the project. County Extension Agents Survey on Expert Systems for Fruit Growers A survey was distributed by electronic mail in January 1989 to measure the familiarity of county extension agents with fruit expert systems and to solicit feedback on the overall expert systems program. The survey was necessary for two reasons: (1) many extension personnel were not informed about expert systems development, thus indicating some training sessions were necessary; and (2) feedback was received that indicated agents in cooperating counties could be better served and utilized by the evaluation process. The survey was sent by PenMail to agents with horticultural responsibilities in all 67 county extension offices in Pennsylvania. Additional questions were asked of agents in the counties where growers were cooperating in the pilot study to solicit feedback on improvements to the evaluation process. A vast majority (84 percent) of county extension agents were at most only somewhat familiar with expert systems for fruit production. Seventy-six percent of agents indicated that they would attend an in-service training program on how to use this technology in their programs. Extension Agent Expert System Training Session In response to feedback from county extension agents, training sessions for county extension personnel were scheduled during the March extension in-service training programs at The Pennsylvania State University. Agents participated in a lecture and discussion of what expert systems are and how they work. In another session, participants received hands-on experience with expert systems in a computer laboratory. This training was provided
OCR for page 423
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS to help familiarize agents with expert systems and to lay the groundwork for the future diffusion of agricultural expert systems. Local Experts Network A proposal has been made to extension administration to initiate a network of extension agents to serve as local experts to support expert systems users within a specified region. The local expert is a person who learns a new technology quickly and is motivated to help others learn it (Landy et al., 1987). Scharer (1983) suggests that the individual is central to the ultimate success of the training effort. This process, which is often used in the diffusion of software technologies, provides a more rapid response to user problems and educational needs than is currently available through Cooperative Extension programs. It is expected that this network will facilitate a more efficient and effective adoption process. CONCLUSIONS The project reported here is the first in the literature of an agriculture-oriented expert systems being tested in the field with comparisons of user and nonuser practices. Evidence from this study supports the thesis of Audirac and Beaulieu (1986) that the access conditions of a technology need to be considered in the diffusion process. Those access conditions of the expert system derived from its technological development as well as its intrinsic characteristics are important variables in the diffusion process. In particular, two characteristics seem noteworthy based on the results of this study. First, the Penn State Apple Orchard Consultant expert system is primarily an information delivery technology. While it contains data base production information (such as weather), it also requires the input of reliable, site-specific information in order to formulate recommendations for the user. The information requested as well as the resultant recommendations require the apple producer to form questions and to look at problems in a manner different from that of previous information delivery systems used in apple production. That this transition will not occur automatically is reflected by the fact that the test group exhibited various levels of use and that almost none of the changes in practices occurred until growers had sufficient time to develop some familiarity with the system's logic. Some growers indicated that they still do not trust the system to make decisions for them. This attitude is appropriate. PSAOC is not intended as a substitute for good management but as a source of information to guide and enlighten growers' decisions. Distrust of the PSAOC expert system could also be the result of incongruence between growers' perceptions of
OCR for page 424
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS the system versus those of their apple orchards. The expert system is an information technology that is intrinsically different from most information technologies that have previously been used by apple producers. The kinds of practical and educational experience a grower or user has may affect how well the system is understood and, thus, adopted. Second, the expert system is a technology that is inherently connected to microcomputers. For a grower to make use of the decision support capabilities of PSAOC, they must (1) have access to a microcomputer capable of running the system and (2) be able to operate the computer proficiently. While the software was designed and developed to be used by people with little or no computer experience, results of the study indicate that growers with the least amounts of computer experience also had the lowest rates of system use. This would appear to be an example of the access conditions of the technology not being congruent with the farming operation. This technology is inherently computer based, and a farming operation must have access to a computer and a person who can operate it before the technology will be adopted. By substituting information for some chemical inputs, the Penn State Apple Orchard Consultant expert system has the potential to contribute to the generation of more sustainable apple production systems in the northeastern United States. This trend can accelerate through the introduction of more information-intensive, low-input IPM practices into the farm production system. This study has provided some preliminary evidence that changes in usual production practices occur as growers and users substitute information for purchased inputs, in this case, pesticides. It was also demonstrated that the substitution of information for inputs was stimulated by the expert system, which enabled the grower or user to collect, integrate, and interpret the information rapidly. However, based on other evidence produced by the study, it appears that the potential for sustainable agriculture that this technology holds will be diminished without some attention to better linking of the access conditions of the technology to the farming operation. RECOMMENDATIONS More work will need to be done at the first stage of the diffusion process if the Penn State Apple Orchard Consultant is to become an effective tool for sustainable agriculture. This first stage concerns the set of activities which provide for the “establishment of diffusion agencies or a network of outlets from which the innovation is distributed to potential adopters” (Audirac and Beaulieu, 1986, p. 63). In the present case, it is planned that this diffusion network will be the traditional Cooperative Extension Service network of university and county
OCR for page 425
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS extension offices and personnel. In addition to acting as the distributive agent for this innovation, this network must also provide new educational training programs in key areas identified by this research, if the effective adoption of this innovation and its potential for sustainable agriculture are to be realized. Some growers are not using the system very often, and others are not being stimulated to change production practices based on their use of the system. In some of these instances, perhaps no change is necessary or advisable. In other instances, change would be highly beneficial in terms of grower profits and reduced pesticide use. In the latter case, effective adoption is not occurring and the potential to reduce the amount of pesticide inputs being used is diminished. To correct this situation when the system is offered for general use by growers, it is recommended that the diffusion agency provide new educational programming in the following areas: training in and basic orientation to computer use for farming operations in general and agricultural expert systems in particular; these training sessions should be held on a very localized basis and taught by people who are familiar with expert systems software and the cropping system being discussed; training that provides an overview of the gradual modification of existing production systems to incorporate reduced-input methods; this training should focus on societal-level needs and responsibilities for reducing pesticide use as well as the long-term farm-level benefits for doing so; establishment of a network of local experts to provide a resource for growers experiencing difficulties with the computer or expert system; continual updating of system capabilities, so that recommendations remain scientifically current and appropriate; training of extension specialists and agents to familiarize them with the possibilities and potentials of the system; and beginning the process by delineating the criteria and goals for sustainable agriculture attainable with expert systems as a tool. In this way scientists will be better able to begin to design production systems for agricultural operations of all sizes that provide more flexibility in responding to dynamic production conditions, thus enabling time and spatially specific recommendations for the expert system to be better implemented. In the long run this may be the greatest contribution of agricultural expert systems development toward a more sustainable system of global agriculture. ACKNOWLEDGMENTS The Penn State Apple Orchard Consultant expert system described here was developed by J. Travis and K. Hickey, Department of Plant Pathology;
OCR for page 426
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS E. Rajotte and L. Hull, Department of Entomology; R. Crassweller, Department of Horticulture; P. Heinemann, Department of Agricultural Engineering; and R. Bankert, V. Esh, J. Kelley, and C. Jung, Integrated Pest Management computer programmers. Program evaluation was conducted by J. McClure and T. Bowser, Department of Entomology; C. Sachs, W. Musser, and D. Laughland, Department of Agricultural Economics and Rural Sociology; and W. Kleiner, Pennsylvania State University Cooperative Extension. Cooperators from other institutions include L. Berkett, Department of Plant Pathology, University of Vermont; D. Cooley, Department of Plant Pathology, University of Massachusetts; and S. Wolfgang, orchard leader, Rodale Research Center. Partial support for this work was provided by LISA project LNE88-8, “Implementation of Electronic Decision Support System for Apple Production.” REFERENCES Audirac, I., and L. J. Beaulieu. 1986. Microcomputers in agriculture: A proposed model to study their diffusion/adoption Rural Sociology 51(1):60–77. Bowser, T. 1990. Adoption of Expert Systems by Apple Growers: A Test of a New Model Unpublished master's thesis. Pennsylvania State University, University Park, Pa. Coulson, R. N., and M. C. Saunders. 1987. Computer-assisted decision-making as applied to entomology. Annual Review of Entomology 32:415–437. Crassweller, R. M., P. H. Heinemann, and E. G. Rajotte. 1989. An expert system for determining apple tree spacing. Hortscience 24(1):148. Denning, P.J. 1986. The science of computing: Expert systems. American Scientist 71:18–20. Heinemann, P. H., E. G. Rajotte, J. W. Travis, and T. Bowser. 1989. An expert system for apple orchard management. Paper presented at the 1989 International Meeting of the American Society of Agricultural Engineers and the Canadian Society of Agricultural Engineering. Hey, J. D. 1979. Uncertainty in Microeconomics. New York: New York University Press. Landy, F. J., H. Rastegary, and S. Motowidlo. 1987. Human-computer interactions in the workplace: Psychosocial aspects of VDT use. In Psychological Issues of Human Computer Interaction in the Work Place Amsterdam: Elsevier/North-Holland Science Publishers B.V. Rajotte, E. G. 1987. A reflective decision support system for Pennsylvania agriculture: Merging electronic information sources, artificial intelligence, and field experience. Agricultural Economics and Rural Sociology Staff Paper No. 144. University Park, Pa.: The Pennsylvania State University. Rajotte, E. G., R. F. Kazmierczak, Jr., G. W. Norton, M. T. Lambur, and W. A. Allen. 1987. The national evaluation of extension integrated pest management (IPM) programs. Virginia Cooperative Extension Service Publication No. 491-010. Blacksburg, Va.: Virginia Cooperative Extension Service. Scharer, L. L. 1983. User training: Less is more. Datamation 175–236.
OCR for page 427
SUSTAINABLE AGRICULTURE RESEARCH AND EDUCATION IN THE FIELD: A PROCEEDINGS Schmisseur, E., and R. Doluschitz. 1987. Expert systems insights: Future decision tools for farm managers. Journal of the American Society of Farm Managers and Rural Appraisers 51(2):51–57. Travis, J., K. Hickey, E. Rajotte, L. Hull, R. Crassweller, R. Bankert, P. Heinemann, V. Esh, and C. Jung. 1990. Penn State Orchard Consultant. University Park, Pa.: The Pennsylvania State University. Wetzstein, M. E., W. N. Musser, D. K. Linder, and G. K. Douse. 1985. An evaluation of integrated pest management with heterogeneous participation Western Journal of Agricultural Economics 10(2):344–353.
Representative terms from entire chapter: