Australian
Agri-Food 2000 Research Forum
Melbourne August 17
An Innovative Approach to Training Agribusiness Consultants
Hargreaves, D.M.G. Hochman, Z. Dalgliesh, N.P. and Poulton, P.L.
CSIRO Tropical
Agriculture /
Agricultural Production Systems Research Unit (APSRU)
PO Box 102 (203 Tor St)
Toowoomba QLD 4350
| Abstract |
| Introduction |
| Background/Context |
| What is FARMSCAPE? |
| Participatory Action Research |
| Instructional Design |
| Summary |
Click
here for the PowerPoint Presentation  |
Farming in Australia is a risky business, and particularly so for the dry-land cropping region of northern Australia. Extreme climatic variability makes it difficult to improve decision making and planning of farm production. The Agricultural Production Systems Research Unit (APSRU) is using Participatory Action Research (PAR) to develop a training program for accreditation and licensing of professional agronomy consultants. In its first phase this program aims to train ten agronomy consultants in a range of tools developed by ASPRU over the past nine years, together known as the FARMSCAPE (Farmers, Advisers, Researchers, Monitoring, Simulation, Communication And Performance Evaluation) approach.
Hassalls & Assoc.; IAMA Limited; Michael Castor & Assoc., and Ward Agriculture are participating in this training program. The FARMSCAPE approach combines the tools of soil and weather monitoring, with sophisticated PC based crop simulation models. These tools are used in a process of joint inquiry between farmers and their advisers to help structure the uncertainty caused by climatic variability.
There were significant challenges posed in designing such a training program: the wide geographic distribution of learners; the high opportunity cost of time consultants’ spend on dedicated off-the-job study; and a wide range of consultants’ professional training backgrounds and experiences.
The FARMSCAPE Training and Accreditation program addresses these challenges with an innovative mixed mode of delivery including printed materials, synchronous net meetings, conference phone calls and face-to-face workshops. Content is delivered primarily ‘on-the-job’ with projects negotiated and designed around consultants’ actual work tasks.
This process was initially designed by APSRU content experts working with a professional instructional designer. By using a Participatory Action Research framework, which formalises a development, deployment and evaluation cycle, the consultants and other stakeholders will become partners in the project design process. It is anticipated that the consultants will play an important role both in the development of the training course and in shaping a new FARMSCAPE approach that is better suited to the consulting profession.
Farming in the dry-land cropping region of northern Australia can be a risky enterprise. Extreme climatic variability makes it difficult to improve decision making and planning of farm production. APSRU is using PAR to develop a training program for accreditation and licensing of professional agronomy consultants. In its first phase this program aims to train ten agronomy consultants in a range of tools developed by ASPRU over the past nine years, together known as the FARMSCAPE approach.
APSRU was formed in 1990 and is a joint research and development unit involving CSIRO Tropical Agriculture and Queenslands’ Departments of Primary Industries and Natural Resources. APSRU’s research focus is on the dry-land farming regions between northern New South Wales and Central Queensland. APSRU’s core technology is the Agricultural Production Systems sIMulator (APSIM) that represents relationships in the production system among crops, soils, weather and management actions, etc. (McCown et al 1996). APSRU also has significant research investments in seasonal climate forecasting, and a range of tools for soil and weather monitoring.
Over nine years of research has resulted in a suite of tools and techniques known as FARMSCAPE. This approach was developed using PAR involving APSRU researchers, farmers and commercial advisers. The FARMSCAPE approach uses the tools of crop simulation, soil and weather monitoring and facilitated group discussion to aid farmers decision making, with respect to a range of production variables, including management of N application and reserves, stored water, crop choice and planting date. The most pertinent finding from this project is that farmers not only found this type of activity relevant - but have indicated a willingness to pay for a service based on this approach.
Due to such demand expressed by participating farmers, our next challenge was to evaluate alternative modes of commercially sustainable delivery. This led to the development and delivery of a Training and Accreditation program for agribusiness. This involves two components: 1) training agribusiness consultants in FARMSCAPE tools and techniques and; 2) joint development of a new FARMSCAPE approach suitable for consultants. The flexibility of using PAR allows each company to develop a system that best suits their clients. We advertised publicly for interested and suitably qualified companies to apply. From the eight companies that applied we selected four: Hassalls & Assoc.; IAMA Limited; Michael Castor & Assoc., and Ward Agriculture.
We have engaged an instructional designer within a PAR framework to help us invent a training and delivery system with our consultant clients. Instructional design is the process of designing and developing an instructional program. Instructional design includes: conducting a needs analysis; designing a system for training; gathering and organising content; designing and producing materials; evaluating suitable modes of delivery, assessment design; and evaluation.
The emergence of the Internet has provided important opportunities for delivering not only training, but also the FARMSCAPE approach to farmer clients. The Internet’s most important contribution, for our purposes, appears to be the increased timeliness and reduction in costs associated with reduced travel that results from meeting online. Secondarily the Internet allows us to develop and deploy interactive training materials with unprecedented efficiency.
The Australian northeastern cropping region is known equally for its highly productive soils, as it is for extreme rainfall variability. Farmers make production investments under great uncertainty and risk due to climatic and market environments. A significant change was also taking place among institutions servicing agriculture. Hochman (2000:2) states the "…public sector extension services were being downsized and highly competitive private sector service industries were developing, especially around the fledgling industry of dryland cotton." This trend suggested that in terms of designing an approach for delivery, we needed to develop a commercially sustainable delivery mechanism, and not one that relied on the rapidly diminishing resources of the public sector extension services.
The FARMSCAPE approach had its genesis in the realisation that despite years of investment, there were virtually no examples of sustained use by farmers of decision support systems (DSS), or the models on which these were based.
"In 1992 a group of APSRU researchers faced the sobering realisation that despite the lapse of 17 years, the lament by Spedding in 1975 that there were "few convincing examples of the use of complex models of agricultural systems being used as a basis for agricultural practice" still seemed valid. The path of model development for ‘systems analysis’ and decision support systems had become well worn, but still had not led to real farms and their management." (Hochman, Z. 2000:1)
FARMSCAPE researchers set out with the modest aim to see if any farmer could value simulation as an aid to management, and under what conditions. From nine years of research has emerged an approach that has had promising results.
The FARMSCAPE approach was developed using PAR as used by Checkland (1985) that is as a framework for developing a methodology. We started out by getting involved with farmers, working with them on problems they saw as important. This initial approach combined the use of "…science-based ‘hard’ systems approach to analysis of problems and the principles of participatory action research." Hochman (2000:2).
Over time it emerged that APSIM played two important roles in enhancing learning and practice. In addition to providing insight into otherwise opaque farming processes (such as the performance of soil water and nitrogen), there emerged a process for aiding farmers’ exploration and development of their own heuristics. The closest analogy for using APSIM in this way is that of a flight simulator.
"If we view learning as a process where an action--> result--> reflection--> learning leads back to further action, flight simulators can facilitate learning by shortening the delay between action --> results. The simulator also demands structural explanations of the action --> result link that will force participants to search for a better understanding of the underlying forces that produce a given set of outcomes. The design of the learning laboratory also increases reflection and enhances learning out of which better decisions can arise" (Bakken 1994 in McCown, R.L. 2000)
FARMSCAPE methodology
This approach would normally involve researchers negotiating joint trials within farmer’s existing commercial crops. This would often involve APSRU researchers measuring farmers’ soil properties, such as available water and nitrogen reserves. In addition to merely providing input data for APSIM simulations, "Soil measurements often produced insights and information that farmers found provided better grounds for conceptualising and managing the soil resources of water and nitrogen." (Hochman, Z. 2000:1)". The data itself provided valuable insights to farmers.
Simulation provided a way for farmers to ask ‘what if’ questions about previous and coming seasons. Often farmers ask, ‘what if I had done this last season’ or ‘how often can I expect this result over the past 100 years". Simulation provides a way for farmers to run ‘virtual experiments’ where experimentation, or learning by trial and error, would be too costly and ineffective.
Traditionally these sessions were held in a face-to-face mode, but with the use of the Internet we are now also able to hold such sessions remotely. RIRDC funded research has enabled us to design a system for researchers to deliver customised simulations via farmers’ home computers, potentially anywhere in Australia.
An outcome of this PAR approach with farmers was the identification of four primary uses for APSIM as part of management decision making. These are: i) benchmarking; ii) production decision support; iii) marketing decision support and; iv) system design and analysis.
BenchmarkingBenchmarking a crop involves running ASPIM for the previous cropping season, with the farmers paddock level soil and weather data. Farmers see the ensuing discussion between predicted potential and actual performance as providing an important opportunity for learning about the performance of their own farms, on a paddock-by-paddock basis.
"Given actual seasonal climate, soil, and management inputs, simulation [can be] used to calculate what a crop should have yielded in the absence of extraneous factors, thus providing a benchmark against which actual crop yield can be assessed." (Hochman, Z. 2000:2)
Case example: Kapun, Queensland.
This example demonstrates a benchmark simulation of a sorghum crop at Kupunn, Queensland. In this case, the simulated and actual yields were very close and so the farmer could conclude that his crop yielded to its potential. Interesting questions still arise: how much less fertiliser nitrogen could have been used without decline in returns? Could higher returns have been achieved with the environmental potential of that season using both a higher plant density and fertiliser rate? How does this change risk as well as returns? (Hochman, Z. 2000)
In benchmarking sessions, such as in the case above, APSIM is used to aid discussions about the ‘gap’ between the predicted potential for last season and the actual outcome. Such discussions are often about such things as: when a crop runs out of soil moisture or nitrogen, when a cold period reduces production at a sensitive growth stage, or the effects when roots "tap into" stored moisture allowing a crop to keep growing through a dry period. Our research demonstrates that such discussion-simulation learning cycles enhanced participants’ understanding of the impact of physical influences on their immediate farming experiences.
Production decision support
This involves farmers using simulation for planning current and upcoming crops. Based on data collected about pre-plant soil water, soil nutrient status, and seasonal climate, simulation has been used to forecast likely outcomes of decisions on crop choice, variety selection, fertiliser rates, sowing date, plant population and row configuration.
"Based on pre-plant soil monitoring data, simulation enabled assessment of expected crop performance in the upcoming season by calculating what would have happened with the same "starting conditions" in past years for which rainfall records exist. Additional skill was added to such forecasts by using the Southern Oscillation Index (SOI) phase system as an indicator of climatic outlook." (Hochman, Z. 2000:2)
Rather than providing a simple prediction of potential future outcomes, this type of analysis and discussion session usually leads to farmers improving, or developing new heuristics. This heuristic may relate to production issues such as sowing window, fertiliser rate or row spacing.
Marketing decision support
Farmers have found APSIM useful for aiding decisions relating to marketing their crops. Farmers that we have worked with have been particularly interested in using in-crop simulation to get a better handle on the potential for forward selling a crop. There is an increasing realisation that it is just as important to sell a crop well, as it is to grow a crop well. Farmers use APSIM to reduce uncertainty around potential yield, which allows them to hedge against likely income fluctuations. Over the past couple of years the price for sorghum has fluctuated between $AUD100/t to $AUD150/t. A grower confident of their likely yield could take advantage of these fluctuation through futures contracts. Farmers in the FARMSCAPE project have used APSIM for in-crop yield predictions, to help with yield estimations for futures contracts.
Case example: Kaimkillenbun, Queensland.
A farmer involved with our research suggested that APSIM could be a useful tool to help provide more accurate prediction of potential yield. He provided us with his soil nitrogen and moisture data for July. APSRU researchers added in the latest SOI figures before running the APSIM model using climate data for the last 50 years.
Simulations were conducted on four occasions: one in July 1998 when soil water and nitrogen data were available, another in December 1998 just after planting when most management parameters (planting date, variety, sowing rate) were known and on two other occasions during the growing season. In each of these simulations the weather data of the past fifty years were used to simulate past outcomes for the remaining portion of the season up to harvest. Results were presented and discussed in terms of probability of outcomes. For example the July simulation indicated a median yield of 5.5 t/ha while yield at the second decile (yield that was exceeded in four out of five years) was 2.75 t/ha.
In July, when prices were high, the farmer forward sold an amount that was equal to the lowest yield outcome in his farming experience. The farmer’s actions regarding the marketing of the 1998-99 sorghum crop did not appear to be influenced by the study until he developed the conceptual framework that would enable him to adjust his worst case yield scenario for such factors as his antecedent stored soil moisture. In the lead up to the 1999 winter cropping season the farmer’s learning from this research became apparent when he forward sold a portion of his wheat crop based on the yield probability indicated by the second decile in the 50-year simulation. The farmer’s explanation of this policy change was that he was still being conservative but he now has a new way of understanding his risk.(Hochman, Z. 2000)
System Design and Analysis
After previous experiences with group discussion aided by APSIM, farmers frequently requested simulations concerning their long-term management strategies. Participating farmers have used APSIM to explore crop rotations over period of up to 100 years. It is important to note that this is not an attempt to find the ‘optimum’ rotation.
Farmers often request a limited number of rotations that they have been considering, to include in the analysis. Farmers often overlay a whole suite of information not adequately dealt with by APSIM, such as management of disease, weed control, their access to equipment and particular financial position.
Typically an initial meeting is held to determine the rotations the farmers wish to examine, and to document their management rules. These rules are used to initialise the simulation, together with relevant rainfall data. Simulations are run and subsequently presented to the farmer group. Often this would lead to a further round of group discussions, where variations of the rotations are run. Farmers were interested in exploring a range of issues, but particularly income per hectare, cash flow and soil loss.
Development of these approaches
These four approaches were developed collaboratively with farmers and their advisers through a Participatory Action Research framework. Our experience is that PAR provides a powerful way to engage key stakeholders in the research process. Below I describe our definition of PAR, how we use it, and a brief history of approaches to intervention in farm management.
Background to our use of PAR
This following section defines PAR for our purposes, and discusses why we use PAR.
We have been active students of the main attempts, over the last century, by professionals to intervene in farming practice. McCown (2000:1) describes a progression of three approaches that are viewed as an evolution; a process of ‘accrual’ and ‘coexistence’ which amount to a significant body of experience. The historical progression of the three approaches part are: i) the development of models that identify the ‘theoretically-optimal action’; ii) development of models to ‘aid farm managers’ procedures’ iii) use of PAR to find better procedures for model facilitated guidance and learning.
Apart from implicit disappointment from the lack of success of the first two attempts, they both provided valuable learning that in turn led to the current interest in PAR.
"The learning that this provides stems largely from the historical succession of attempts in which failure of an approach gave rise to an improved approach which in turn failed, but from which important learning occurred. Disappointing non-success in practice is offset by the progression in ideas that constitute growth in understanding of the enterprise of ‘decision support’ and point the direction for a more promising model-assisted intervention practice in the future" (McCown, R.L. 2000:5)
PAR based intervention: how do we define and use PAR?
We used a PAR approach directly involving Farmers and Advisers to jointly develop the FARMSCAPE approach.
Our definition of PAR draws heavily on Checklands’ (1981) description:
The concept of action research arrises in the behavioural sciences and is obviously applicable to the examination of human activity systems carried out through the process of attempting to solve problems. The core is that the researcher does not remain an observer outside the subject of investigation but becomes a participant in the relevant human group. The researcher becomes a participant in the action, and the process of change itself becomes the subject of research. In action research the roles of researcher and subject are obviously not fixed: the roles of subject and practitioner are sometimes switched: the subjects become researchers… and researchers become men (people) of action (Checkand, P. 1981:152)
…using this methodology is akin to constructing an ‘ad hoc theory’ about the problem situation which is derived ‘neither from general theory nor from scientific testing, but from a pre-formed set of concepts developed in experience’ (Checkland, P. 1981:251)
The process involves: i) developing a methodology using PAR ii) using this methodology in real world problem situations; iii) being reflective about its use; iv) and making subsequent modifications based on the reflective process. We are again using PAR to jointly develop a methodology for using the tools and techniques of the FARMSCAPE approach within the agribusiness industry.
Instructional DesignIn order to effectively aid the design of a ‘new FARMSCAPE approach’ with agribusiness consultants, we considered it important to first transfer the tools and techniques that make up FARMSCAPE to those advisers.
We engaged an instructional design expert within a PAR framework, to work with us on developing an appropriate plan for instruction. This plan essentially included a training needs analysis, material design, designing a mode of delivery, trainee assessment and process evaluation.
Background to our approachIn a broader context we are mindful of recent developments in the ‘theory-of-learning’, as Billett states (1998: 264) "…views about learning are being re-appraised to consider how the circumstances of the acquisition of knowledge influence cognition and, as a consequence the transfer of knowledge to other situations."
The emphasis throughout this training is to provide ‘situated’ learning experiences to trainees. Billett (1998:263) defines situated learning as "…learning through goal directed activity situated in circumstances which are authentic, in terms of the intended application of the learnt knowledge."
Training needs analysis
As a first step we conducted a training needs analysis to identify significant characteristics of the client group. The consultants are from four organisations, Hassalls & Assoc.; IAMA Limited; Michael Castor & Assoc., and Ward Agriculture. They are distributed geographically between central Queensland to northern New South Wales. Learners generally have tertiary qualifications; a background in rural science; and requirements for training to be both timely and efficient. None of the participants can afford to spend significant portions of time in off-the-job learning activities. There is likely to be significant variation in participants previous experience; nature of their work; and learning needs.
These consultants work full-time, through visiting farms and providing advice to clients face-to-face, via telephone and fax, and increasingly though the use of the Internet. Advisers typically reside in a branch, which services a specific geographic region. All branches have ready access to computers and are currently being upgraded to high-speed Internet access. Training is encouraged and is regarded positively by management.
Module content
The training course has been divided into six modules, and is designed to be delivered in a progression, from module one through six. Built into this are two levels of accreditation. Level one accreditation will be conferred upon successful completion of modules one and two. Level two accreditation will be conferred at the completion of the remaining modules three through six.
These modules are:
1. Soil monitoring and data management: principles, techniques, and quality assurance;
2. Weather monitoring and data management: principles, techniques, and quality assurance;
3. APSIM: the program and the science;
4. Simulation applications in farm management;
5. Analysis of simulation results and quality assurance;
6. Flexible representation of results and communication with decision-makers.
Module development methodology
Each module has a module coordinator, who provides leadership on the content. This person is a research scientist who writes the content for the module. Each module is reviewed by a panel of scientific peers who essentially ‘referee’ the content. The module coordinator makes changes before submission for formatting and layout.
The central content of each module is a training manual, with a set of associated materials. The manual includes descriptive content, examples, case studies and exercises. Secondary materials may include web-based content, such as online questionaries, online lectures or tutorials.
Mode of Delivery
We are using a flexible approach to delivery, which includes traditional face-to-face workshops, printed materials, interactive and static web based content and online net meetings. The training needs analysis informed the process of designing the mix of delivery modes. We have adopted an innovative mode of delivery, and one that larger more traditional educational providers are still coming to terms with.
During the early stages of design, the Internet was seen as central to providing flexibility in delivery of learning materials. The Internet provides both an efficient delivery mechanism for materials in a traditional format, and the development of new materials which take advantage of the interactive possibilities the Internet offers. The Internet also offers possibilities for supplementing face-to-face interaction with new forms of computer mediated interaction, which substantially reduce the cost associated with physical distance. The Internet supports several modes of interaction including: i) synchronous computer-mediated-communication between facilitators and learners; ii) asynchronous computer-mediated-communication between facilitators and learners; iii) and on-line interactive courseware.
We are trialing using Microsoft Media Technologies for delivering online lectures including voice, video, and animated slides / applications. We have successfully used Microsoft NetMeeting™ for synchronous Internet workshops between candidates. Initial evaluation of these sessions has proved remarkably positive. There is a wide gap between candidates’ competence with computers, so we have experienced initial challenges in connecting all participants in a timely way. However, after this initial period, the meeting progressed in a remarkably smooth way. We’ve had ten sites participating for up to two hours with minimal disruption.
Assessment
The primary mode of trainee assessment is by electronic journal. The electronic journal provides the central component of the trainees action learning cycles. The journal encourages a process of reflection on current activities. Module tutors are included in this loop and are therefore in a position to aid learner’s reflective process and to provide guidance and to initiate corrective action where necessary. Site visits by module tutors also provide an important source of information on progress. Assessment is largely progressive and is competency based. Accreditation is conditioned on demonstration of stated competencies in all modules.
Our research has indicated that farmers find value in an approach to aid production related risk management in highly uncertain climates, known as FARMSCAPE. Demand for FARMSCAPE tools and techniques has grown to such a degree among farmers in northern New South Wales and Central Queensland that APSRU could no longer meet that demand in a sustainable way.
Faced with continued disinvestment by successive governments in the public extension network, FARMSCAPE researchers looked to the private consulting industry. This led us to develop a Training and Accreditation program, which has the dual aims of transferring FARMSCAPE tools and techniques, and secondly inventing a new FARMSCAPE through a PAR framework directly with consultants and their clients.
As this paper is published candidates, having completed modules one and two, are approaching level one accreditation. Modules three through six will be delivered over the coming period, and we look forward to reporting on our progress and learning’s at a future time.
References
Billett, S. 1998. Situated Learning: bridging sociocultural and cognitive theorising. Griffith University, Australia.
Checkland, P. 1981. Systems Thinking, Systems Practice. John Wiley & Sons, Chichester.
Hochman, Z. et al 2000. Case Study: The FAMRSCAPE Experience in INRA: Learning and knowing processes in agriculture in industrialised countries. (in print)
McCown, R.L. 2000. Learning to Bridge the Gap Between Designed Decision Support and Farming Practice: A systems analysis of model-based interventions in farm management. (in print)
McCown, R.L. 2000. Learning to Bridge the Gap Between Designed Decision Support and Farming Practice: Learning from the failure of intervention which used theory-based recommendations for practice. (in print)
