Previous PageTable Of ContentsNext Page

Understanding adoption of precision agriculture technologies

C. Ambrosio, C. Linehan and G. Kaine.

Department of Primary Industries –Tatura Centre, Victoria, Australia. E-mail cinzia.ambrosio@dpi.vic.gov.au

Abstract

As agricultural practices can potentially have severe environmental impacts on our landscapes, the wider community is demanding that agricultural producers change their practices to ensure environmental sustainability for future generations. In response, the Victorian grains industry is investing in Precision Agriculture (PA) innovations, such as Variable Rate and Site Specific Technology (VR and SST), to minimise nutrient loss and seepage and hence reduce environmental impacts, while sustaining production levels. VR and SST may provide growers with the capability to apply nitrogen variably in a paddock, ie. delivering appropriate amounts of nitrogen across a paddock thereby optimising response to fertiliser, minimising deep losses and reducing the cost of fertiliser applied. However, despite the predicted potential benefits, farmer adoption rate for VR and SST has been low.

In this study we use personal interviews underlined by a theoretical framework based on the integration of Consumer Behaviour Theory and Farming Systems Theory frameworks to understand the factors influencing adoption of VR and SST. We found that growers invested substantial time to determine the application rates for nitrogen fertiliser. There were a number of factors farmers used to determine nitrogen application rates, including; weather forecasts, crop and paddock history and previous cropping experience. The majority of growers had used yield maps to assist them to determine nitrogen application rates. However their experience with these tools had proven to be of a confirmatory nature, reaffirming what they already knew. Regardless of the type of tool used, all growers we interviewed went through a thought out process to determine nitrogen rates. We found that growers applied the same quantity of nitrogen within a paddock but they often varied application rates between paddocks. Growers believed that variability within their paddocks was not high enough to warrant specialised, micro-level treatment within paddocks. These results indicate that, in current circumstances, grain growers did not perceive that VR and SST provided growers with benefits above what they were achieving with their current farming practices.

Key Learnings: (1) The application of nitrogen fertiliser is highly involving because of its critical importance to a profitable and sustainable crop. (2) Grain growers do not deem variability within a paddock to be sufficient enough to warrant more precise treatment. (3) In the current climate, growers perceive that VR and SST do not supply benefits above current practices and therefore adoption is likely to be low.

Key words

Agronomic decision-making, nitrogen fertiliser, needs, risk, informed choices

Introduction

The Victorian cereal industry accounts for nearly 15% of the gross value of agricultural production and in 2000 grain exports contributed $1.2billion towards the state’s economy and represented 8.2% of Australia’s total grain production (Department of Primary Industries, 2004a). To ensure the continued prosperity and sustainability of the cereal industry, the Department of Primary Industries (DPI) and industry partners invest significant resources in research, development and extension to enable on-farm practice change (Department of Primary Industries, 2004b).

An example of current DPI and industry investment concerns work on Precision Agricultural (PA) technologies. PA technologies are management tools intended to assist growers manage their farm productivity at a precise, tailored level. PA technologies consist of a suite of technologies which may include Global Positioning Systems (GPS), yield monitors and maps, auto steer tractors, Controlled Traffic (CTF) systems, and Variable Rate and Site Specific Technologies (VR and SST). PA tools can be adopted individually or as a package.

VR and SST can allow growers to apply fertiliser at variable rates within a field. To be able to use VR and SST, grain growers may need to invest in one or more items of equipment, which can range from basic yield maps, Geographical Information Systems (GIS) and computers, to more sophisticated and costly systems, which include real-time crop sensors and imaging systems. For the purpose of this study, the term VR and SST refers to highly sophisticated equipment that includes real-time crop sensors such as thermal/in-crop hyperspectral imaging systems.

Real time crop sensors detect early stress signals and are able to assess nitrogen status in cereal crops. With this capability, inputs can be applied according to crop requirements, increasing the efficiency and effectiveness of nitrogen use. It is suggested that the adoption of VR and SST will lead to nitrogen fertiliser savings for growers. However, despite assumed benefits of VR and SST, adoption has been low.

Conversely other PA technologies, such as CTF and autosteers have achieved greater adoption compared to VR and SST. CTF permits growers to direct their machinery on predetermined, established tracks allowing for subsequent passes occurring over the same pathway. This means that only some areas within a paddock become compacted whilst the linear tracks provide tractors with guidance for consistent, precision application of inputs. Autosteer tractors rely on GPS to accurately determine where the machinery is within a paddock. These capabilities allow growers to reduce overlap such as seeding rates and chemical applications.

In this paper we propose to use the integrated theoretical frameworks developed by Kaine (2004) to assist us to understand the reasons for the apparent low adoption rate of VR and SST among grain growers. In addition, we propose the findings of this study can be used to determine where future investment in VR and SST research and development could be directed to potentially achieve greater adoption of VR and SST.

Background

Assael (1998) states that a key factor regarding the purchase of products is the time and effort, or level of involvement, a consumer will spend before purchasing a product. If the product is unimportant to the consumer then involvement will be low. Low involvement products tend to be purchased often, are inexpensive and generally involve minimal risk for the consumer. If the product is important to the consumer then involvement will be high, with the consumer putting significant effort into evaluating the product prior to purchase. Key attributes that important or highly involving products tend to display include goods or services that are:

  • rarely purchased,
  • costly and;
  • financially and socially risky.

Kaine (2004) suggests changing farming practices can be ‘highly involving’ because such changes exhibit the aforementioned attributes of highly involving products. For example, changes in grazing management or the planting of a new crop variety have the potential to substantially impact on farmers’ incomes and, in many instances, their lifestyle. Furthermore, the integration of a new practice or technology into an existing farming system generally requires careful planning and management.

Kaine (2004) argues that in circumstances of evaluating potential adoption of innovations that are important to them, farmers tend to follow complex decision making processes; spending time and effort to ensure any change they implement better meets their need. Therefore decisions to adopt or not adopt are based upon an informed evaluative process.

This view of adoption implies that farmers will actively seek information about new practices they perceive as relevant to their needs. Sometimes the information search and adoption process can extend over several years (Kaine and Niall 2001). Secondly, this view implies farmers are unlikely to retain apparently outdated or inefficient practices simply through ignorance, tradition or conservatism. Thirdly, this view implies that the decision not to adopt a new practice will be founded on a reasoned argument. Consequently, attempts to promote adoption of innovations that do not address such arguments are likely to meet with limited, if any, success (Kaine et al 2005).

Methods

Study area

The study was conducted in grain growing regions located in north and southwest Victoria, Australia. A total of twenty-two interviews were carried out. Six interviews were in the Mallee Catchment Management Authority (CMA), ten in the Wimmera CMA, three in the Glenelg-Hopkins CMA and the remaining three in the Corangamite CMA (Figure 1).

Figure 1. Distribution of the catchment areas studied across the State of Victoria, Australia.

Study approach

As growers have spent time and effort to formulate criteria to assess whether any change they are considering will better meet their needs, they are able to logically explain their decisions regarding changes in farming practices including the adoption of new technology (Kaine 2004). We therefore used market research techniques based on Consumer Behaviour Theory and Farming Systems Theory to assist us in understanding their decision making process regarding current practices of nitrogen application as it relates to adoption of VR and SST (Kaine et al 2005).

Study methods

Our research methods approach was consistent with the Kaine et al (2005) framework. We used convergent (Dick 1998) and laddering interviewing techniques to understand the factors growers used to decide on nitrogen application rates and the factors influencing adoption of PA technologies. Convergent interviewing techniques are useful as they allow for the identification of similar and dissimilar patterns in reasoning on a topic. In addition, this technique enables the identification of the reasons for these patterns. We also employed laddering techniques (Grunert and Grunert 1995) to systematically explore the reasoning underlying the decisions and actions of the interviewee. Laddering allows the interviewer to understand the perceptions and reasons behind a given response, oftentimes by simply asking a series of ‘why’ questions. The power of this interview process lies in identifying common and complementary patterns of reasoning among interviewees (Kaine 2004).

To ensure that we interviewed farmers from a variety of farming systems, growers from different regions in Victoria with varying farm sizes were sought, using snowballing sampling (Sarantakos 1998). Interviewing continued until recurring patterns and themes - irrespective of location and/or size - was achieved. Individual interviewee responses were recorded manually by two interviewers, with the notes later transcribed. The notes were analysed using case and cross-case analysis (Patton 1990). Case analysis occurs with the investigation of each individual interview or ‘case’ for accuracy, both in terms of required data and actual transcription. Cross case analysis occurs when all interview data is grouped according to recurring patterns or themes.

Results

Nitrogen fertiliser application decision-making process

We found growers in all regions suggested that nitrogen fertiliser, alongside chemicals for controlling pest and disease, represented the most substantial costs in production of cereal crops. We also found that despite regional differences in the amount of nitrogen applied annually to cereal crops, the criteria growers used to determine nitrogen application rates and timing were similar across all regions. Therefore, for the purpose of this paper, the term ‘growers’ refers to respondents across all the studied areas.

Growers believed the timing of the nitrogen application and the amount applied at each application was critical for maximising yield. Growers suggested that the purpose of the first nitrogen application is to supply enough nitrogen for the plant’s establishment phase. Growers stated that the second application was applied at the end of tillering, that is, when lateral shoots have developed from the axillary bud of leaves at the base of the stem. The second application is designed to aid grain filling, which begins when the plant ceases its vegetative growth stage and diverts nutrients to produce grain seeds. In most cases growers did not apply the annual nitrogen requirement in a single application. Rather, nitrogen was applied on two or three occasions at particular times of the growing season. The first application was typically applied either two to three weeks before sowing and/or at sowing. The second application was applied at the end of tillering when lateral shoots have developed from the axillary bud of leaves at the base of the stem.

All the growers interviewed believed they were applying the appropriate amount of nitrogen to their cereal crops. Growers justified this belief on the basis that they obtained continuing increased yields. They believed that factors other than incorrect nitrogen applications where responsible in years were production was lower than expected. All growers suggested that applying insufficient nitrogen fertiliser would lead to significant reductions in crop yield potential. In addition, they believed that applying too much nitrogen fertiliser would not be as detrimental to crop yields compared to applying insufficient nitrogen. Consequently, growers were more likely to apply higher rates of nitrogen than lower rates.

Growers used various information sources to aid their decision-making for nitrogen fertiliser application, including consultants, seed and fertiliser company representatives and government extension officers. Growers considered a number of factors when deciding their nitrogen application rates. The critical factors growers considered were the paddock’s history (eg. fertiliser used, previous crops sown, current crop and future crop) soil type, previous productivity, soil fertility tests and current and predicted rainfall. In some cases yield maps were used but oftentimes these were informing growers of issues that growers already knew of. In other instances yield mapping was not practical for reasons such as poor signal and/or technological drop out due to insufficient satellite reception or cloud cover. Differences in these factors between paddocks usually meant growers applied nitrogen at different rates in different paddocks. Differences between paddocks in nitrogen fertiliser applied were typically between 10-40%. While growers were confident that different application rates per paddock were warranted, the majority of growers interviewed did not consider it worthwhile to vary application rates within a paddock even if growers were aware of existing variability of different degrees.

We apply nitrogen as a blanket operation [within paddocks]. We rely on the paddock’s history and seasonal conditions to guide us in terms of nitrogen application. If we were recommended to use different amounts of nitrogen we would need to know why.

Factors influencing adoption of VR and SST

None of the growers interviewed in this study had adopted VR and SST. However, growers in all regions demonstrated, to varying degrees, knowledge of VR and SST. This knowledge had been gained by seeing on-farm demonstrations, attending field days, reading about the topic and/or from conversations held with other growers and consultants. Growers were aware that VR and SST was being developed to aid growers in determining nitrogen application rates. However, most growers did not believe VR and SST provided them with significant benefits above their current method of determining nitrogen fertiliser application rates. The reasons for this varied between growers. For example, some growers did not believe the variability within their paddocks was high enough to justify the cost of VR and SST. Other growers were not convinced that the technology was reliable enough due to past experience with other PA technologies. Other growers thought VR and SST was more likely to err on the side of fertiliser application rates compared to their existent practices, which growers perceived would reduce the crop yield. Therefore the risks associated with VR and SST adoption were perceived to be much greater than their current application methods. In this climate of uncertainty, VR and SST was not considered reliable enough to warrant adoption.

We have paddock variability but it is very limited…. I do not manage it [within paddocks] differently for fertiliser or nitrogen application. There is not enough variation to consider it [VRT and SST] seriously being of benefit.

Had problems loading it [PA technology] up to the computer. It’s only new technology. Might consider it [PA technology] once all the bugs are sorted out. Need to iron bugs out first.

More [nitrogen] on rather than less…..You’ll never go broke buying fertiliser.

Adopted PA technologies

During interviews growers related their experiences with other PA technologies that had been adopted. For example, boom sprays with light bars, autosteer tractors and CTF were mentioned throughout the interviews as innovations that had either been adopted by the interviewee, or by another grower known to the interviewee. Growers gave a variety of reasons for adopting these innovations. For example, boom sprays and autosteer had improved application accuracy of inputs such as, chemicals or seeds, thereby significantly decreasing costs, time and effort. The adoption of raised bed systems had accelerated the uptake of CTF in some areas. This was because raised beds had eased major problems of waterlogging in the Glenelg-Hopkins and Corangamite areas, which are prone to substantial rainfall and poor drainage issues. In these areas, cropping was uncommon years ago prior to the adoption of raised beds.

CTF – it sold itself to us…. We have proof that it has contributed to [improved] yields. It has helped with waterlogging, is more efficient, has improved soil health and it makes everything a lot easier management wise.

Discussion

Our results demonstrate that innovations that aim to change nitrogen applications such as VR and SST are highly ‘involving’. This finding is consistent with Kaine (2004) and Kaine et al. (2005). This means that growers will invest substantial time and effort to ensure that any innovation that alters nitrogen application does not put their farming business at any risk. Therefore, proposing any changes to nitrogen application would need to provide growers with a surety that they would obtain better results compared to current methods, which growers do not currently believe is the case with VR and SST.

Whilst growers believed applying different nitrogen application rates per paddock is warranted, they do not believe applying varied rates within a paddock is necessary. Growers believe they have either insufficient variability or VR and SST will increase the risk of under fertilising a paddock, which will decrease crop yield. Furthermore, growers believe the technology required to use VR and SST is not reliable enough to base critical decisions on, such as nitrogen application. Consequently we believe that the decision not to adopt VR and SST must be due to a mismatch of grower’s current needs and the benefits provided by the innovation.

Conversely, growers had adopted some PA technologies, such as CTF, as they believed these had provided a benefit above their previous practices. A prime reason for the adoption of CTF was the introduction of raised bed cropping systems in southwest Victoria which had changed their farm system. This allowed for the production of profitable cereal crops in areas where grain growing would have been inappropriate previously. Hence CTF suited their changed farming system. In all cases, growers were able to explain why one PA technology met a need whilst others did not.

We do not infer that grower’s perceptions and/or practices cannot be improved. Rather, if there are substantial demonstrable benefits to be gained from PA adoption, including VR and SST then we believe that further investment needs to occur in research and extension activities. Adoption is likely to be low until the benefits to be gained clearly exceed current practices and risk factors are reduced.

Conclusion

Consistent with the framework of Kaine (2004), we believe that VR and SST that aims to alter important farming decisions, such as nitrogen application, will be highly involving and invoke complex decision-making. We found growers currently apply nitrogen at a uniform rate within paddocks but variably between paddocks. Growers do not see a need to alter this practice as they deem VR and SST as being too risky compared to what they are currently doing. In the current climate, growers do not see soil or yield variability within paddocks as a big enough ‘issue’ to warrant adoption of VR and SST. This implies that the adoption rates of VR and SST in the current farming system will continue to be low. However, PA technologies, such as CTF, had been adopted as these provided growers with benefits above current practices and had fulfilled a need.

Given the critical role of nitrogen, we suggest credible information be provided on the extent and nature of within paddock variability so that farmers and extension agents may consider what actions (PA or other) may be appropriate in developing more profitable and sustainable systems. In areas of high variability, where it can be demonstrated that production is suffering due to this, on-farm trials might need to continue so as to counter this perception of insufficient variability. If VR and SST have the ability to provide greater benefits compared to current methods, then efforts need to be made to highlight and demonstrate these benefits - from a growers perspective - through various extension programs.

References

Assael H (1998). Consumer behaviour and marketing action, South-Western College Publishing, Cincinnati, Ohio.

Department of Primary Industries (2004a). Grains Industry Strategy. http://www.dpi.vic.gov.au/dpi/nrenfa.nsf/FID/-A9C8D320E60D0FD14A2568E000189391?OpenDocument.

Department of Primary Industries (2004b). Corporate Plan 2004 -2007. Department of Primary Industries, Victoria.

Dick B (1998). Convergent interviewing: a technique for qualitative data collection. http://www.scu.edu.au/schools/gcm/ar/arp/iview.html.

Grunert K and Grunert S (1995), ‘Measuring subjective meaning structures by the laddering method: theoretical considerations and methodological problems’. International Journal of Research Marketing, vol.12, pp209-225.

Kaine G, Bewsell D, Boland A and Linehan C (2005). ‘Using market research to understand the adoption of irrigation management strategies in the stone and pome fruit industry’. Australian Journal of Experimental Agriculture, vol. 45, pp 1181-1187.

Kaine G (2004). Consumer Behaviour as a theory of innovation adoption in agriculture. Social Research Working Paper 01/04. AgResearch Ltd, New Zealand.

Kaine G and Niall E (2001). Sheep breeding: complex decision making and brand loyalty. Report to the Victorian Department of Natural Resources and Environment, School of Marketing and Management, University of New England, Armidale, NSW, Australia.

Linehan C (2001). Increasing the impact of research and extension in irrigated agriculture. Milestone Report, Department of Primary Industries, Tatura, Victoria

Linehan C and Kaine G (2003). The key to reducing the gap between research and application. Paper presented at the National Landcare Conference, Darwin, Northern Territory, Australia.

Patton MQ (1990). Qualitative evaluation and research methods, 2nd ed, Sage Publications: USA p158.

Sarantakos S (1998). Social research. 2nd edn, Macmillan, Victoria, Australia.

Previous PageTop Of PageNext Page