Adoption and Use of Precision Agriculture in Brazil : Perception of Growers and Service Dealership

Precision agriculture (PA) is growing considerably in Brazil. However, there is a lack of information regarding to PA adoption and use in the country. This study sought to: (i) investigate the perception of growers and service dealership about PA technologies; (ii) identify constraints to PA adoption; (iii) obtain information that might be useful to motivate producers and agronomists to use PA technologies in the crop production systems. A web-based survey approach method was used to collect data from farmers and services dealership involved with PA in several crop production regions of Brazil. We found that the growth of PA was linked to the agronomic and economic gains observed in the field; however, in some situations, the producers still can not measure the real PA impact in producer system. Economic aspects coupled with the difficulty to use of software and equipment proportioned by the lack of technical training of field teams, may be the main factors limiting the PA expansion in many producing regions of Brazil. Precision agriculture work carried out by dealership in Brazil is quite recent. The most services offered is gridding soil sampling, field mapping for lime and fertilizer application at variable rate. Many producers already have PA equipment loaded on their machines, but little explored, also restricting to fertilizers and lime application. Looking at the currently existing technologies and services offered by dealership, the PA use in Brazil could be better exploited, and therefore, a more rational use of non-renewable resources.


Introduction
Precision agriculture (PA) involves the development and adoption of some techniques to improve the management of agricultural systems, aiming to optimize inputs applications such as fertilizers, pesticides, seeds and irrigation resources to reduce inputs costs and maximize the crop production (Bora et al., 2012), besides to reduce environmental impacts (Bramley et al., 2008).In several crop production Brazilian regions, PA has been played an important role in crop production systems, mainly due to the technical and economic benefits that PA provides over the years.Costa and Guilhoto (2011) stressed that the benefits of PA adoption, impacts directly on social and economic benefits of Brazilian agricultural economy.However, the benefic effects of PA are more restricted to large cropped areas, usually operated by major companies linked to crop production.Pierpaoli et al. (2013) found that the size of cropped area is the most important parameter to farmers when they have to decide to adopt PA, due to the higher possibility to increase income.Then, properties with large cropped area has more potential to be capable to invest large amount of resources, time and learning in order to use PA technologies compared to properties with small cropped area (Adrian et al., 2005).reduction of fuel consumption and machine operation time.Besides the economical aspects, negative environmental impacts arising should be reduced with PA adoption due to a more rational use of inputs in crop production systems (Bramley et al., 2008).Australian farmer's point of view about the low adoption of PA in Australia relays on technology frustration and the lack of technical support in the field.They pointed out that technology costs was not the overriding factor to PA adoption (Mandel et al., 2010).On the other hand, Batte and Arnholt (2003) analyzed six farms in Ohio (US) that recently adopted PA technologies and the profitability was the major factor that motivated farmers to adopt PA, although even not all farmers surveyed have observed the global profitability linked to PA adoption.The farmers surveyed also pointed out that on-farm research, quality information generated by PA to support decision and risk reduction in the environmental contamination were the major concerns to adopt PA.More than a decade ago Swinton and Lowenberg-Deboer (2001) concluded that PA adoption/expansion would increase slowly in area with high population and, in area with less cropland available unless the environmental benefits would be very well reported.Reichardt and Jürgens (2009) found in German conditions that some major issues related to PA adoption were lack of technical support to PA tools and lack of knowledge to manage correctly the data to apply them correctly in the crop production system.For German farmers, the systems incompatibility among several companies' suppliers stills the major constraint to PA adoption.Batte and Arnhorld (2003) concluded that to increase PA adoption, the development of more simple technologies is the most important contribution to support farmers in the decision making process.In Alabama, farmers PA adoption is related to well establish farmers with large cropped areas and more educated level (Adrian et al., 2005).
Precision agriculture adoption survey has been conducted over the years in United States.In 2011, 85% of the respondents reported that they have been used at least one PA technology in the crop production system (Whipker & Erickson, 2011, 2013).Similar results were found in the previous survey in the same region (Whipker & Akridge, 2009).Although similarity among results was found regarding to PA technology use, Whipker and Erickson (2013) found that the use of GPS guided systems with autocontrol/autosteer was the major used over years.This fact emphasizes that implementation of new PA technologies is more suitable to be embraced for farmers that already use any PA technology available.For instance, PA technology adoption by new users has been increasing annually with 76 and 83% in 2007 and 2008 respectively (Whipker and Akridge 2007;2008).
There is little information available regarding to PA adoption in Brazil.Silva et al. (2011) demonstrated that PA adoption in sugarcane production increased sugarcane yield and quality, and also increased profitability to farmers.On the point of view of sugarcane industry, the reduction in the environmental impact was the major issue to take attention.The use of PA technologies will be essential to sustainability of Brazilian agribusiness and mainly to achieve higher crop yield while reduce environmental impact (Silva et al., 2011).Precision technology adoption use can affect directly the economy at regional and large scale (Costa & Guilhoto, 2013).A study that evaluated scenarios such as i) increase in crop yield; ii) input reduction; iii) increase in crop yield and reduction in inputs, and; iv) increase in crop yield and increase in inputs, concluded that the major impact was on increase of crop yield, that impact directly in social benefits (employment raised) and economics benefits (increase of income) to Brazilian economy (Costa & Guilhoto, 2013).The benefits of inputs reductions is solely to increase farmers income and it is not reflect in economic benefits to the society; then, the benefits to the society must be analyzed on the point of view of reduction to environmental impact (Costa & Guilhoto, 2013).Silva et al. (2007) attempting to clarifying the costs of PA technology in order to increase the adoption by farmers, carried out a comparative analysis of the costs and economic profitability involved in implementing PA and conventional farming practices in the state of Mato Grosso do Sul.Even though PA presented higher effective operational cost comparing to traditional agriculture due to technical assistance, maintenance of sophisticated equipment, yield and soil mapping, for example, the unitary cost (i.e., the cost per kilogram) in the precision agriculture system was lower than the cost in the tradictional system.
The benefits of PA adoption are widely known and transferred to agricultural systems.Although crop yield is one the major factor that impact PA adoption in some cases, is important to scientist to understand how the perception of farmers and agronomists about PA are in the crop production systems.Research with farmers, agronomists and users about their perceptions of PA technologies are limited in Brazil and there is a gap of information and knowledge that must be filled.Then, our study sought to investigate the perception of growers and service dealership about PA technologies and to identify constraints to PA adoption and also to obtain information that might be useful to motivate producers and agronomists to use PA technologies in the crop production systems.

Method
A web-based survey was used to collect data from farmers and services dealership involved with precision agriculture in several crop production regions of Brazil.Survey used was similar found in Silva et al. (2011) to evaluate PA adoption in Sao Paulo State (Brazil) to sugarcane production system.
We previously identified users of PA technologies to obtain their perceptions and support needed and followed the same approach of Diekmann and Batte (2010) adjusted for Brazilian conditions.After this process, we develop a web-based survey to obtain information from two groups: 1) farmers that use AP technology, and; 2) professionals and companies that are technical support providers or farmer consultants.The division in two groups was the same approach used by Reichardt and Jürgens (2009).Web-based survey was developed based on the same approach in Whipker and Akridge (2009) adapted to Brazilian conditions.Both web-based surveyes were accessed by respondents with the follow links: https://sites.google.com/site/agriculturadeprecisaotocantins/which was developed to farmers to answer it; https://docs.google.com/spreadsheet/viewform?formkey=dG9UUG 51WEhtTVZQSXJCR0ZNTVRna3c6MQ to be answered by professionals and companies that are technical support providers or farmer consultants.The sampling frame used to select the respondents was lists of individuals from Precision Agriculture Network coordinated by Embrapa (Brazilian Agricultural Research Corporation) and stakeholders, machinery companies' technical representatives and crop production sales personnel.Similar approach to sampling respondents was used to Larson et al. (2008) and Dieckman and Batte (2010).For each group separately an e-mail was sent them explaining the survey goals and the respective link to the web-survey.Based on prior published data that involved survey (mail or web), we expected about one third of the invited to respond the web-survey (Larson et al., 2008;Dieckman & Batte, 2010).

Web-Survey to Farmers
The survey included questions about respondent's as follow: year started PA in the farm; total area used with PA considering lime and fertilizers application with variable rate; use PA in the decision making process for crop and soil management; soil sampling using grid or GPS; soil sampling grid size; soil sampling in soil layers; how PA is adopted in the farm (own equipment or contract the service); list of PA equipment's used in the farm; farming operation used as variable rate application (VRA); observations related to PA use (e.g.crop production cost reduction); problems found regarding to equipment's maintenance and software support; PA technical support; investments; observations in increase crop yield; constraints to adopt PA at farm and regional scale.

Web-Survey to Professionals and Companies that Are Technical Support Providers or Farmer Consultants
The survey included the following questions: major company activity; how long the company is on the PA market; average size of the farms assisted; PA market grow since started to work with PA; total area in hectares the service dealership assist; PA area in the farms assisted; service most required by farmers; average soil sampling grid size; percent of PA service in the company income; PA impact on crop production reduction costs to assisted farmers; PA equipment's available to be used in the farmers that are assisted; company grow expectation in the next years; expectations to increase PA adoption at regional scale and the constraints observed and found to consolidate PA as crop and soil management practice to increase nutrient use efficiency.

Data Analysis
Primary data obtained from the web-survey were analyzed considering percentages of questions answered (Silva et al., 2011).A total of 250 e-mails were sent to farmers and 10% were answered from several states such as Goiás (GO), Rio Grande do Sul (RS), Paraná (PR), Maranhão (MA) and Tocantins (TO).Although returned answers was relatively low, it was similar that was found in similar studies with survey research (Whipker & Akridge, 2009;Holland et al., 2013;Watcharaanantapong et al., 2014;Whipker & Erickson, 2013).The answer from professionals and companies that are technical support providers or farmer consultants reached a large number of crop production region.The respondent companies were from states of São Paulo (SP), Mato Grosso (MT), Bahia (BA), Paraná (PR) and Rio Grande do Sul (RS).Answers from companies were from several regions of Brazil despite the company headquarter was located in one of the listed states above, giving more accurate results.

Vision of Farmers that Are PA Users
Analyzing farmer respondents we identified that soybean and maize are the most common crops that PA is used (Table 1).Both crops represent the major Brazilian commodities.Silva et al. (2007) already demonstrated that PA technology may guarantee higher production, decreasing the unitary cost and, consequently, making the system more rewarding on a long-term basis for soybean and maize.On the other hand, the adoption of PA      Regarding to service provides, farmer's respondents agree that it is too difficult to keep updated with constant new technologies in PA, since dealership provides the PA service or the option to purchase of machinery, but did provide free technology upgrades.Furthermore, respondents also agree that data collection can suffer climatic or operational interference that were not fully corrected by software, thus compromising data accuracy and therefore the recommendation.The lack of PA monitoring for dealerships related in updating and maintenance of software and equipment, training and technical accompaniments, make the field teams remain limited on proper handling of equipment.Pieces for timely replacement are also important obstacles.Note.1: Completely disagree; 2: Partially disagree; 3: Fully agree; 4: Partially agree; 5: Not agree/not disagree.
Although more than half want to invest less than 10% of the previous year's income (Figure 6b) in PA tools, hindering great adhesion to PA technologies and still be many limitations to overcome (Table 3), farmers expect improvements in some precision agriculture technologies, or new PA technology options (Figure 7).Farmers have great expectation arising new tools for recommendation and application of fertilizer and lime.In the same scale of expectations (78%), an integrated interpretation considering data analysis in different database is expected.With a lower frequency of responses, but expected still with 67 percent, the variable rate seeding,   s did not know ed that three-q a) and for the strategy to con of the PA ma y have, since be to perform thi ees (Figure 3).still very inci  We can also notice that the PA branch companies focus their efforts on providing services rather than the sale or rental of PA products.This may be related to the fact that large companies of machinery and equipment now offer this technology at the moment of the machinery sale.This information can be a big opportunity for companies offer a rental service to small farmers who plan to take PA technologies on their property, since they do not need to purchase new machines with load technology.This is another option for service providers expand its operations.
Dealership also responded one series of questions about the limitations for increasing the adhesion to PA technologies.Although a small portion of farmers reported that they agreed that the producer cost was higher than the benefits (Table 3), the vast majority of companies and service providers reported that totally disagree that the costs are higher than the benefits (Table 5).This lower perception of farmers can be due to many of them do not have a detailed control of the cost with and without application of PA tools.
Comparing the results from Table 3 and Table 5, we obserced similarity between the answers given by producers and service providers.However, some points should be highlighted.For example, around 45% of farmers respondent figured out benefits of PA adoption in the own business (Table 3).This percentage could be higher according to the service providers' vision, or even among those who realized this benefit it could be more significant.Once the companies responded that 75% completely disagree that their customers who could have benefits from PA technologies are already users.While 25 percent fully agree (Table 5).There is negotiation on the costs of services provided to farmers 0 0 75 25 0 Note. 1: Completely disagree; 2: Partially disagree; 3: Fully agree; 4: Partially agree; 5: Not agree/not disagree.
A point that is worth emphasized is that according to the producers, one limitation is finding replacement spare parts (Table 3).However, the service providers as a whole disagree with this statement, reporting not having trouble keeping replacement parts in stock (Table 5).
Services providers companies were also surveyed about expectation of new PA technologies (Figure 12).In general they showed lower expectations when compared with producers.However, it is noteworthy that a higher expectation regarding to technologies for implementing the variable rate seeding.Such service also have been reported in the last USA survey, being among the services that more is expected to grow by 2018 (Widmar & Erickson, 2015). jas.ccsenet.

Figure 12
While mo provides r between th in the rang PA activity Results sh questionna many opp needs som crop produ PA adoptio 2001).

Conclus
The

Figure
Figure Figure 9.

Table 2
trained to operate machines and equipment with embedded PA systems, contribute for low investment in PA or new users over the years. %

Table 3 .
Limitations that constraint PA adoption to increase responded from farmers from several Brazilian regions in the crop season 2011/2012.Values in percent (%) calculated from answers signed in each specific item

Table 4 .
Products/services offered by companies that providing services and trends.Values in percentages calculated from the responses marked on each item.Data from questionnaires answered by dealership in the crop season 2011/12 %

Table 5 .
Limitations that constraint PA adoption to increase responded from dealerships from several Brazilian regions in the crop season 2011/2012.Values in percent (%) calculated from answers signed in each specific item.