=Paper=
{{Paper
|id=Vol-1152/paper40
|storemode=property
|title=Precision Agriculture Applications In Horticultural Crops In Greece and Worldwide
|pdfUrl=https://ceur-ws.org/Vol-1152/paper40.pdf
|volume=Vol-1152
|dblpUrl=https://dblp.org/rec/conf/haicta/GemtosFA11
}}
==Precision Agriculture Applications In Horticultural Crops In Greece and Worldwide==
https://ceur-ws.org/Vol-1152/paper40.pdf
Precision Agriculture Applications in Horticultural
Crops in Greece and Worldwide
Theofanis A. Gemtos, Spyros Fountas, Katerina Aggelopoulou
Laboratory of Farm Mechanisation, University of Thessaly, Greece, gemtos@agr.uth.gr
Abstract. Precision agriculture is the management of spatial and temporal variability of the
fields using ICT. The application in horticultural crops was developed in the last ten years.
Data are collected from different sources (yield and quality, soil properties, remote sensing),
stored to GIS data bases, analysed using geostatistical methods to develop management zones
and decision support systems are used to assist farmers to the management. Variable rate
application systems were developed to apply inputs according to the real requirements of the
plants in the management zones. Although variability is a well established fact, farmers’
adoption is rather slower than expected because the system is complicated and in many cases
profitability is not well demonstrated. Additionally, environmental benefits cannot take direct
monetary values for the farmer. PA has a wider application impact like precise movement and
management of farm machinery, development of efficient mechanization and fleet
management as well development of farm management information systems.
Key works: Precision agriculture, horticultural crops, decision support systems, adoption,
and profitability.
1. Introduction
Precision agriculture (PA) can be defined as the management of spatial and temporal
(within the growing period and between the years) variability of the fields using
Information and Communications Technologies (ICT). PA was in some way applied
in the past. The small farm manager was able to observe all variation within the
fields and take appropriate decisions for each part. Mechanisation and increase of
farm size reduced this knowledge. The larger the field, the lower the knowledge of
the variability. The average rule was used to manage the fields. When the first yield
monitors were developed it was proved that yield and soil properties varied highly
within a field. This fact marked the development of precision agriculture.PA is
aiming at increasing farmers’ knowledge of his field and return better management
based on this new knowledge. PA has a rather short history. Its application started
about 20 years ago when GPS and new sensor technologies were made available
(Heraud and Lange 2009). The initial applications were mainly for arable crops. The
applications in horticultural crops were rather delayed and started by the end of the
1990’s and the last decade. PA is a cyclic system of data collection, use for the crop
management, evaluation of the decisions and the cycle continues for the subsequent
years (Figure 1). Each year data are stored in a data base (library) and is used as
historical data for the future decisions.
The objectives of this paper is to give an account of the progress made the last few
years in application of precision agriculture in horticultural crops as a mean to
enhance its use by the Greek agronomists and farmers.
2. Data Collection
451
Many types of data can be collected during the growing season. All have to be
geo-referenced using GPS technology and introduced to a GIS data base.
2.1 Yield mapping
Yield mapping can be carried out easily in mechanized crops. In vines sensors were
developed relatively early for the mechanical harvesting of grapes for wine making.
They were applied in 1999 vintage in Australia and in the USA (Arno 2009). They
used either loading cells that weighed the crop passing on a conveying belt or an
array of sonic beam mounted over the grape discharge chute to estimate the volume,
and the tonnage, of fruit harvested (Bramley and Hamilton 2004). The results in
Australia showed 8-10 fold difference of yield between parts of the same parcel
(Bramley 2001a). Temporal variability is an important factor in the development of
stable management zones. Research in arable crops (cereals Blackmore et al.2003
and in cotton Fountas et al. 2004 ) showed that the trends after the third year are
cancelling out and we can only define areas of stable high and low yielding and
unstable yielding. Tree crops seem to have more stable yields as Bramley and
Hamilton (2004) and Bramley et al. (2003) found in vineyards after five years data.
Figure 1. Presentation of a precision agriculture Figure 2. Yield mapping in
system (Markinos et al. 2002) apples in Greece.
In handpicked fruits yield mapping is more difficult. In Florida’s citrus plantations,
Shueller et al.(1999) used a system to weigh the palette bins where the oranges were
collected. The bins were removed by a hydraulic lift which used loading cells to
weigh them and a GPS to record the position. It was assumed that each bin
represented the yield of the surrounding trees. Yield variability was observed in a 3.6
ha orchard. In Greece Aggelopoulou et al.(2010a) mapped the yield in apple
orchards. The apples were handpicked and placed in 20 kg bins along the rows of the
palmette shaped trees (Figure 2). Each bin was weighed and geo-referenced. The bins
corresponding to 5 or 10 trees were grouped to represent their yield. A similar
approach was used by Tagarakis et al. (2010) for yield mapping of vines. Yield
spatial variability was evident in all applications even in orchards of 1 ha. Fountas et
al. (2011) measured the yield variation in olive trees orchard. Olives, in conventional
orchards, were picked by hitting the fruit branches by sticks. Olives were falling on
plastic sheets underneath each tree. The olives were placed in bags and left in groups
for loading to a platform. Each bag was weighed and geo-referenced using a GPS.
Each group of bags was considered to present the yield of the surrounding trees and
was the basis for the yield map. Spatial variability was also present. Ampatzidis et
al.(2009) have mapped the yield of peaches. They used RFID or bar code tags on the
452
bins. A weighing machine was combined with a tag reader and a GPS to record the
weight and the place of each bin. The data collected was used to produce yield maps
of the orchard. Konoatski et al. (2009) have mapped the yield of a 1.6 ha pear
orchard. They measure the yield of each tree (harvested in three passes).
Qiao et al. (2005) developed a mobile automatic grading robot. It was moved from
plant to plant. Workers picked the peppers and placed them on the machine for
grading. The machine located the plant, weighed the fruits of each plant and analysed
the quality. Yield and quality maps showed spatial variability even in the very small
plot of the experiment. Akdemir et al. (2005) have measured the yield variability in
dry onions in Turkey. They have divided the field in 10X10 m grid and they had
collected onions from each grid by hand and weighed them. They found a yield
variation from 10 to 50 t/ha
2.2 Quality mapping
Quantity and quality are the two components of the field production. Quality is very
important and its variability was the object of relevant research. Several laboratories
are working to develop sensors to measure quality of products. In high value crops
quality offer premium prices and increased income to the farmer.
Extensive work on the grapes’ quality was carried out. Grape samples were taken and
analysed to assess its variability. Using remote sensing they found high correlation
between the vegetation indices maps near veraison (beginning of maturity) and the
grape quality maps. Based on that, they separated the production of the two zones
which produced different quality of wines. The dense vegetation part gave lower
quality with lighter colour (Bramley et al. 2003). But it was not always true that low
yielding parts produced high quality (Bramley and Hamilton 2004). Bramley (2005)
has studied the quality variability in commercial fields. Quality variation was there
but was much lower than yield’s. The zones formed by the quality parameters were
not always similar to the yield zones. He concluded that it is difficult to define zones
of certain quality characteristics as the wine industry is requiring. Cost of samples
collection and analysis is high and only on the go sensors could offer the opportunity
to separate qualities of grapes. Best et al. (2005) measured an index m2leaf/kg-fruit in
vines. They found that quality of grape (Brix, colour) were lower when the index was
larger (higher vigour of the plants). Sethuramasamyraja et al. (2010) used a hand held
NIR spectrometer to analyse anthocyanin variability in two vineyards for two years
in CA, USA. The vines were divided into two management zones based on threshold
values suggested by the vineries. A harvester with two stores was developed and
used. Different quality grapes from the two zones were directed to the appropriate
store. The two quality lots produced different quality wine and proved the usefulness
of the method. Aggelopoulou et al. (2010a) have analysed the spatial variability of
quality of apples. They measured several parameters of the quality like colour,
sugars, malic acid, pH and flesh firmness and found negative correlation between
yield and quality. The variability existed even in small size orchards.
2.3 Soil properties analysis
Soil is the substrate where crops are grown. It affects several parameters of crop
growth, the final yield and its quality. Most of the cropping activities are also
affecting soil through tillage, compaction fertilization etc. Soils were analysed for
453
their properties. Grid sampling of different size was used. A parcel size of 0.4 ha was
considered reasonable for commercial applications. Samples taken from the parcel
were mixed, homogenised and then analysed for their properties. Soil maps were
produced for each property and could be used to define fertilization. Fountas et
al.(2011) using a grid sampling and analysis of an olive orchard defined the soil
maps (Figure 3) and the amount of P and K fertilization for each tree.
P K
13,2 - 98,2 ppm 193,0 - 680,0 ppm
51.3 591
34.3 394
57.4 408 344 404 406
73.2 39.4 49.4 540 435
72.1 41.4 31.2 441 261
98.2 37.0275
37.0275 286 369
45.1 529 337 342
51.4 22.3 43.2 440 347 302
48.4 33.2 382
51.6 31.2 58.3
380 310 347 216
52.2 35.1 52.4 38 345 337 460 449 378
40.4 62.4 36.2 36.9 40.2 37.027
37.027 589
406 427 285 488 405
75.4 412 412 207
72.4 82.3 36.2 45.6 40.4
45.2 60.2 53.3 276 581
364 395 474
357 510 521 320
56.8 58.8 65.2 37.0265 305
46.4 78.3 28.7 298
52.6 62.3 70.2 62.3 90 286
37.0265 352
51.3 343 365 208 388
380
Lat
48.4 31.4 306
Lat
52.3 29.2 38.1 29.6
60.6 542 315 226 350
80.2 37.026 295 388
343 271
46.2 31.4 16.5
37.026 68.3 43.4 224 680 210
80.6 26.2 273 402
328 228
69.2 77.4 38.2
55.4 26.2 37.0255 228 296 378
39.1 41.2 45 338 334
580
37.0255 44.2 46.4 33.2 218 328
52.1 13.2 196 595 306
41.3
33.2 31.2 37.025 208
464
42.3 45.2 34.9 360 248
37.025 16.2 0
28.7 448 304 305
36.2 0 435
56.8
37.0245 193
250
40.3 70.2 29 21.631 21.6315 21.632 21.6325 21.633 21.6335 21.634
41.2
37.0245 40.3
34.2
21.631 21.6315 21.632 21.6325 21.633 21.6335 21.634
Log
Log
Figure 3. P map of olive K map of olive orchard Figure 4. Prescription
orchard map for N application
per group of trees.
Aggelopoulou et al. (2010) have analysed soils in a dense grid. They found that
correlations between soil nutrients and yield was not consistent. They suggested
taking into account apples’ yield and the nutrients removed to produce prescription
maps for fertilizers application. Best el al. (2005) found also low correlation between
soil properties and yield parameters but better between yield and ECa maps.
Soil sampling and analysis is labour intensive and costly activity. For research
purposes this can be justified but not in most commercial applications. A second
possibility is to define management zones with another measurement like yield and
direct the soil sampling to the zones. This highly reduces the samples and the cost. A
third possibility is to develop sensors that can measure soil properties on the go. This
is a fast and usually low cost method. Several methods were developed or are under
development. The soil sensors were based on electrical and electromagnetic, optical
and radiometric, mechanical, acoustic, pneumatic, and electrochemical measurements
(Adumchuk et al.2004). Electrical resistivity and electromagnetic induction (EM)
measure soil apparent electrical conductivity (ECa). This property is directly
connected to soil properties like texture, water content, organic matter, salinity, ions
in the soil and temperature. If we exclude saline soils and take measurements near
field capacity then measurements are correlated to soil texture. Many researchers
have reported correlation of yield and ECa (Kitchen et al. 2005). Soil texture is a
basic factor of soil variability and influences several soil and crop parameters.
Heavier or lighter soils react differently to weather; require different water, fertiliser
and herbicides applications. The GPS readings when they are relatively accurate can
offer at the same time elevation maps.
Aerial and satellite images obtained using remote sensing help in analyzing the
variability of soil (Adumchuck et al. 2004). Soil colour without vegetation offers an
indication of its texture and soil organic matter. Early laboratory studies showed
correlation of soil OM with both visible and near infrared (NIR) reflectance.
454
Mechanical sensors have been used to assess soil compaction using instrumented
tines (Andrale et al. 2002) or automatic penetrometers. They gave good results but
they have to pass through the soil to assess compaction. Electromechanical sensors
have been developed. One with commercial application can map pH. Soil samples
were taken on the go and electrodes can measure pH.
2.4 Remote sensing
Remote sensing is the group of techniques than can collect field data without being in
contact to plants, soil etc. An electromagnetic wave when falling on an object it can
pass through, reflected or absorbed. Measuring these effects we can have useful
information. It is a useful technology for PA as it can give data for parameters of the
field relatively easily. In general we see the reflected sun light that is formed by the
ultraviolet wave lengths, the visible light (Red, Green and Blue) and the infrared.
The green plants are absorbing the red and blue wave lengths and reflect the green
and the infrared. Measuring the reflected wavelengths with a multispectral cameras
we can measure the vigour of the plants or any problem like disease, nutrient
deficiency or water logging etc. We can correlate soil colour to the organic matter,
moisture etc. Light reflectance (sun or some artificial light source) has been used in
PA in the form of vegetation indices. The most used of them is the Normalised
Vegetation Index (NDVI). Several other indices can be calculated and used offering
good agreement with certain characteristics of the crop. NDVI has been correlated to
crop yield and quality. The measurements of plant reflectance can be carried out by
satellites, airplanes or ground instruments.
In several PA studies crop reflectance was used as an early measurement of the crop
growth and for prediction of yield and product quality. Bramley et al. (2003) have
used NDVI of vines at vaireson as an indication of grapes quality and used it to
separate the product into high and low wine quality producing lots. The idea was
successful and gave a good results and a profit to the farmer (see later). Best et al.
(2005) in Chile, found good agreement between NDVI and yield and quality of a
vineyard (correlation coefficient r2>0.7) and between LAI and NDVI (r2>0.75). Hall
et al. (2010) studied the correlations between spectral images and the properties of
the grapes and yield. They estimated canopy area and canopy density, which were
consistently significantly correlated to fruit anthocyanin and phenolic content, berry
size and yield. But total soluble solids correlations were not stable.
Any object when have a temperature above absolute zero emits electromagnetic
radiation. This is used in thermal cameras that can detect differences in temperature
in plants. Thermal cameras have been used in precision agriculture to assess water
status of crops and regulate irrigation (Agam et al. 2009).
3. Data analysis and management zones delineation
All data collected have to be analysed and interpreted. Simple exploratory
(descriptive) statics can give a first idea on the values, their spread, the range and the
distribution. Geostatistics, are used for spatial interpolation. Final construction of the
thematic maps for successive years are made using spatial variability structure of the
sampled data (variogram) and an interpolation method (kriging). Semi-variograms
are used to assess the spatial variability of the measured values. Maps covering the
whole field can be produced and indicate the variability of the properties. There are
455
several methods of data analysis although that there is not a clear method to compare
the produced maps. We are still based on optical impression for the comparison of
the maps. Correlations between parts of the field with different parameters can be
carried out to assess their relationships. Kitchen et al. (2005) tried to delineate
productivity management zones based on ECa, elevation and yield using MZA. They
used a pixel agreement between zones to compare the zones based on different
parameters. Taylor et al. (2007) have presented a protocol for data analysis and
management zones delineation using available free software to help farmers in the
better use of the data collected through precision agriculture technologies.
Soft computing techniques have been employed to define correlation between the
properties measured and permit a forecast of the results. Neural networks, fuzzy
logic, fuzzy cognitive maps has been used recently to analyse data and explain yield
variation. (Papageorgiou et al. 2010).
The analysis of the data aims at defining parts of the field called management zones
with common characteristics which can be managed in a common way. Management
zones delineation should form homogeneous parts of the field where inputs or other
practices can be applied in the same way. The management zones should be large
enough to permit VRA of inputs but small enough to be homogeneous. Management
zone delineation can be done using fuzzy cluster analysis.
4. Variable Rate Application (VRA)
VRA technology is the major target for PA. All information gathered should result in
a better management of the formed zones. VR means that the appropriate rates of
inputs will be applied leading either to reduced inputs, costs and environmental
effects or improved yields and quality. Two methods are used to apply VR. The first
called map based, is based on historical data (previous or present year). Process
control technologies allow information drawn from the GIS (prescription maps) to
adjust fertilizer application, seeding rates, and pesticide selection and application
rate, thus providing for the proper management of the inputs. The second, named
sensor based, uses sensors that can adjust the applications rates on the go. The
sensors detect some characteristics of the crop or soil and adjust the application
equipment. VRA can be applied to all inputs. Both systems have advantages and
disadvantages. The on the go sensors are more acceptable by the farmers. Probably
using a mixture of both will offer most advantages in the future.
Variable fertilizer applications in vineyard could help minimizing variability in vine
growth as well as fruit quality (Sethuramasamyraja et al. 2010). Devenport et al.
(2002) applied VR fertiliser in a vineyard for four years. They have analysed the
nutrient content of the soil and concluded that N and K applications benefited the
field as they reduced variation but not the P application where the CV remained high.
Based on management zone delineation and historical data prescription maps can be
produced defining the specific requirements of each zone. The prescription map is
imported to the controller of the application machine and changes the adjustment (the
amount of the input applied per unit of area as prescribed) as the machine moves
through the field. Obviously a lot of data have to be collected and properly analysed
to make effective the application. In tree crops where temporal variability is lower
this application is more feasible than in arable crops.
456
Prescription maps can be produced based on several characteristics of the field or the
crop. In the case of the orchard of Fig. 3 (Fountas et al.2011) the farmer applied the
fertilizer by hand in each tree. He was able to use the map with the two zones and
apply one or two portions of fertilizer in the defined trees. In apples Aggelopoulou et
al. (2010b) have used the soil analysis data and the nutrients removal from the soil by
the crop to prepare prescription maps for fertilizer application (Fig 4). Prescription
maps can be based on characteristics measured during the growing season.
Aggelopoulou at al. (2011) found high correlation between estimated variability of
flowers and yield distribution. This can be used to manage the inputs of the crop as
low yielding parts requirements are different that high yielding.
Several on the go sensors have been presented and used. The most known is the
sensor that detects light reflectance from the crop. Using NDVI the sensor detects the
vigour of the crop. Usually crops with sufficient nitrogen are greener than plants with
lower nitrogen. This characteristic was used to adjust N rates in crops like cereals. In
tree crops several characteristics can be used to directly adjust inputs. Tree canopy
volume, density and height can be measured electronically (Giles et al. 1988). In
citrus orchards of Florida, tree canopy measured by ultrasonic or laser sensors was
correlated to yield. This property was used to adjust the variable chemical
application. (Giles et al. 1988, Turbo et al. 2002). Pulse width modulation nozzles
that use fast reaction solenoids to open or close the flow several times per second can
be used to vary discharge. One other idea changes the active ingredient solution by
introducing it at different rates in the distribution tubes of the sprayer (after the
pump). (Ess and Morgan 2003). Gil et al. (2007) tested a variable rate application
sprayer in vines. The sprayer had three nozzles groups in each part of the row.
Ultrasonic sensors were sensing the canopy width and adjusted the sprayer achieving
58.8% savings.
Variable rate irrigation is of great importance due to the shortage of water reserves
and the importance of irrigated crops in many parts of the word. Applications in
central pivot systems using prescription maps based on soil properties, crop
conditions and the real conditions of the field proved that considerable saving in
water and energy can be achieved (Perry et al. 2002). In a feasibility study of fields
in Greece and Turkey based on soil variability savings of up to 7% (range 2.5-7.2) of
water and energy can be achieved (Gemtos et al. 2010). Perry and Milton (2007)
estimated 12% water savings while Hedley et al. (2009) at 7%.
In orchards, irrigation systems have to be designed from the beginning to achieve
variable rate irrigation. Knowing the soil variability it is possible to develop more
than one networks applying different water depths or frequency of application. The
zones separation criteria are soil texture and soil elevation. Wireless systems of
sensors were developed to measure soil water content during the growing season.
The sensors can give information to the farmer or directly to the controllers of
automatic irrigation systems that can define proper application levels.
Several direct sensing systems have been used for weed control. Some herbicides are
sensitive to soil organic matter. Soil organic matter detection was used to
automatically adjust the herbicide application rate. Increased efficiency was reported
(lit). A second line of action is the detection of green plants and use herbicides only
where the weeds are. The system is to be used between the rows of vegetables or
other crops. More than 30% herbicide savings were reported. In the same line weed
457
recognition systems can be used and drops of herbicides are applied only on the
weeds. These systems work also on the crop row. High herbicide savings are reported
(Bak and Jacobsen 2003). A third line of action is the use of mechanical weed control
by avoiding crop plants. The system defines the crop plants position. There are two
ways. One to detect the seed placement in the field using a RTK-GPS and then
produce maps with plants’ position. The second is to use a camera in front of the
machine to detect crop plants and direct a tool only to the weeds. Several tools were
developed. The most successful commercially is a horizontal disk system that has
one sector removed (Figure 5). The machine vision system or the plant map or both
detect the crop plants and adjust the discs rotation in such way to avoid damaging
them (Dedoussis et al. 2007)
5. Decision Support Systems for the Farmer
A decision support system (DSS) is a computer-based system that supports business
decisions. In agriculture it refers to the decision taken by the farmer for the
management of the farm. Precision Agriculture is directly connected to decision
making by the farmer. PA (Viticulture) can be basically described as an example of
the conversion of data into decisions (McBratney and Whelan, 2001). It is quite true
that research is not successful in that respect at the moment. The lack of functional
tools for decision-taking, explains to certain extend the difficulty faced so far for a
rapid and widespread adoption of PA. This is a fact recognized by researchers in the
field. Arno et al. (2009) pointed that the development of Decision-Support Systems
(DSS) in PV undoubtedly remains a pending assignment. Kitchen et al. (2005)
pointed that more precise crop models working in PA can help in the development of
successful DSS.
Figure 5. Curved disc Figure 5. Selective weed Figure On the row
with one sector removed control machine prototype weed control withtines
6. Profitability and Adoption of Precision Farming
The adoption of a new technology by the farmers is a difficult procedure. The
evolution of agriculture in many parts of the world resulted in aged farmers and
usually of lower education level. This makes changes and adoption of new
technologies even more difficult. Different surveys indicate a lower use of computers
and internet by farmers. Kuter et al. (2011) defined farmers’ adoption of PA as the
combined utilization of several site-specific technologies using GPS such as auto
guidance and VRT of inputs and/or yield mapping on farm. The farmers to adopt a
new system have to recognize, research, and implement these technologies and
458
management practices at an on-farm production level (Kosh and Khosla 2003).
Kuter et al. (2011) pointed that farmers will adopt PA when they are sure that offers
economic benefit, advantages over traditional methods and it is less complicated.
Additionally farmers like to observe an application and see the benefits before
adopting any innovative technology. Research showed that large farms and young
farmers adopt more PA. Ehsani et al. (2010) have reported that the farmers expect
from the new technologies to be proven and robust, cost effective and when new
equipment will be employed to be reliable and well backed up for service and repair.
Early and accurate yield predictions are important.
Adoption is wider in the USA. In 2003, 32% of Ohio farmers had used one PA
component and this percentage increased from previous studies. Larger farms
showed larger application rates (Batte et al. 2003). In Europe adoption is rather low.
It is wider in the North than in the South. Wider to arable than in horticultural crops.
A lot of small farms in Europe make adoption difficult. It is suggested that
cooperative use of equipment or through contractors can help to that direction.
The economic returns of PA have been studied. It is clear that PA application
requires some new equipment (yield sensors, installation of equipment, ECa sensors,
VRA equipment, computers etc) that has to be depreciated. Additional costs for
training are also required. Variable costs are the every year data analysis and
interpretation. All these costs should be covered by the benefits from the application.
In many cases improved yields and reduced costs are the benefits and can be directly
estimated. Reduction of chemicals, water or energy use, which apart from the direct
reduction of costs have additional benefits to the environment is difficult to be
translated in monetary units. In high value crops quality improvement can be of great
interest. Bramley et al. (2003) in a separate harvest of the two parts of a vineyard the
high quality grapes gave wine of high price ($30/bottle) while the low quality low
price wine ($19/bottle). They comment that if the grapes were harvested all in bulk
they would produce low quality wine. The profit based on the gross price of wine
was around $30,000/ha. An estimation of the application cost was at $11/t of
harvested fruit which is negligible compared to the profit.
7. ICT in Agriculture
Precision agriculture is not only site specific management. Most of the
technologies used in precision agriculture can be used in several applications
improving farm management. The use of GPS technology can offer guidance systems
to the tractors that help them to follow desired paths in the field. This can lead to
more accurate tree planting or controlled traffic in vegetable fields. The addition of
GPS and other sensors to the tractor (using the ISO BUS standardisation) can offer a
full record of the farm machinery movement as well as fuel and energy consumption.
Recording of farm machinery activities (with inputs form the farmer) can lead to
Farm Management Information System that can cover administration requirements
for certification of production systems (like integrated crop production management
systems) or EU cross compliance (Future Farm 2011). Keeping records on inputs and
yields we form the first step of a traceability system so required by the consumers.
Knowing the machinery movements we can estimate better use or better itineraries
that can improve efficiency. This can save time and fuel but also reduce soil
compaction. The development of autonomous vehicles can led to improved
459
mechanization systems with fleets of small sized tractors working 24 hours a day and
doing accurately all farming activities (Blackmore et al. 2007).
8. Conclusions
From the presented data, it can be concluded that:
Yield, quality and soil spatial variability is present in most of the fields, even in
small size. Therefore precision agriculture can benefit the farmers.
Data analysis still requires better methods especially in correlating thematic
maps. Decision support systems have to be developed to enhance PA adoption.
VR technology has developed methods for site specific applications that can not
only reduce costs but have additional beneficial effects to the environment.
PA has a wider impact in farm management through more efficient machinery
management.
References
Adamchuk V.I., J.W. Hummel, M.T. Morgan, S.K. Upadhyaya (2004) On-the-go soil
sensors for precision agriculture Comp.& Electr. in Agriculture 44 (2004) 71–91
Agam N, A. Ben-Gal, Y. Cohen, V. Alchanatis (2009) Optimal Time-of-Day for
Thermal Remote Sensing of Water Stress in Olive Orchards. ASA- CSSA-SSA 2009
annual meeting, Pitsburgh, PA
Aggelopoulou K.D., D. Wulfsohn, S. Fountas, T.A. Gemtos, G.D. Nanos, S.
Blackmore (2010a)Spatial variation in yield and quality in a small apple orchard
Prec. Agriculture (11), 538-556
Aggelopoulou K. D, D. Pateras, S. Fountas, T. A. Gemtos and G. D. Nanos (2010b)
Soil spatial variability and site-specific fertilization maps in an apple orchard Prec.
Agric.(12) 118-129
!!"#$%$&#$&' ()*' +)' ,$-./01*' 2)' ($&/1$1/3/.01*' 4)' 5$&6/31*' 7)' 8"9/$1*' 8)' :36$1'
(2011). Flower spatial variability in an apple orchard. Prec. Agriculture 12:118–129
Akdemir B.,K Belliturk, C. B. Sisman, S. Blackmore (2005) Spatial distribution in a
dry onion field (a precision farming application in turkey) J. Central European
Agriculture 6(3) 211-222
Ampatzidis Y.G., Vougioukas S.G., Bochtis D.D. and Tsatsarelis C.A., 2009. "A
yield mapping system for hand-harvested fruits based on RFID and GPS location
technologies: field testing". Precision Agriculture 10(1), 63-72.
Andrade, P., Upadhyaya, S.K., Jenkins, B.M., Filho, A.G.S., 2002. Evaluation of UC
Davis compaction profile sensor. Paper 02-1185, ASAE, St. Joseph, MI USA
Arnó J., J. A. Martínez-Casasnovas, M. Ribes-Dasi and J. R. Rosell(2009) Review.
Precision Viticulture. Research topics, challenges and opportunities in site-specific
vineyard management Spanish J. of Agricultural Research 7(4), 779-790
Bak T. Jakobsen H. (2003) Agricultural Robotics Platform with four wheel steering
for weed detection Biosystems Engng 87(2) 125-136
Batte M.T. , C. Pohlman, D. L. Forster and B. Sohngen (2003) Adoption and Use of
Precision Farming Technologies: Results of a 2003 Survey of Ohio Farmers OSU
AED Economics (AEDE-RP-OO39-O3).
Best S., L. León and M. Claret (2005) Use of Precision Viticulture Tools to Optimize
the Harvest of High Quality Grapes FRUTIC 05, September, Montpellier France
460
Blackmore, S., Godwin, R., Fountas, S., (2003). The analysis of spatial and temporal
trends in yield map data over six years. Biosystems Engng (2003), 84 (4), 455-466.
Blackmore, S., Griepentrog, H. W., Fountas, S., Gemtos, T., 2007. Specifications for
an autonomous crop mechanization system. Agricultural Engineering International:
the CIGR Ejournal. Manuscript PM 06 032. Vol. IX. September, 2007.
Bramley R.G.V., 2001a. Progress in the development of precision viticulture –
Variation in yield, quality and soilproperties in contrasting australian vineyards [on
line].Available in http://www.crcv.com.au/research/programs/one/bramley1.pdf
Bramley R (2001b) Precision Viticulture - Research Supporting the developemtn of
optimum resource management for grapes and wine production Available in
http://www.crcv.com.au/research/programs/one/workshop14.pdf [March 2011].
Bramley R., Bruce Pearse, Phil Chamberlain (2003) Being profitable precisely - a
case study of precision viticulture from Margaret River The Australian & New
Zealand Grapegrower & Winemaker Annual Technical Issue 2003
Bramley R and R.P. Hamilton (2004) Understanding variability in winegrape
production systems. 1. Within vineyard variation in yield over several vintages.
Australian J. of Grape and Wine Research 10, 32–45
Davenport J.R., J. M. Marden, and L.Mills (2002) Variable Rate Fertilizers for Grape
Nutrient Management.http://www.grapesociety.org/Meeting%20Proceedings/2002/
fertilizer. pdf
Dedousis A. P. (2007), An Investigation into the Design of Precision Weeding
Mechanisms for Inter and Intra-row Weed Control, PhD Thesis, Cranfield University
https://dspace.lib.cranfield.ac.uk/bitstream/1826/2009/1/Dedousis%20PhD%20Thesi
s.pdf
;.1360' <)*' 4)' 436=3>36*' 36?' @)' +093' ABCDCE' Grower Expectations of New
Technologies for Applications in Precision Horticulture. Florida Cooperative
Extension Service, University of Florida. http://edis.ifas.ufl.edu
Ess D.R., Morgan M.T. (2003) The precision farming guide for agriculturalists, Third
Edition, Deere and Company, Moline, IL, USA
Fountas S., S. Blackmore, T. Gemtos, T. Markinos (2004) Trend yield maps in
Greece and the UK 2nd HAICTA Conf. Proceedings. Vol 2 p 309-319 Thessaloniki
Fountas S., K. Aggelopoulou, C. Bouloulis, G. Nanos, D. Wulfsohn, T. Gemtos, A.
Paraskevopoulos, M. Galanis (2001), Site-specific management in olive tree
plantation. Prec. Agriculture 12:179–195.
Gemtos T.A., B. Akdemir, U. Turker, G Mitev (2010) A Feasibility Study for
Variable Rate Irrigation in the Black Sea region: Economical and Environmental
Benefits. Final Report, Black Sea Cooperation
Gil E., Escola A.,Roseli J.R.,Planas S.,Val L.(2007) Variable rate application of plant
protection products in vineyard using ultrasonic sensors. Crop Prot 26(8)1287-1297
Giles, D. K., M. J. Delwiche, and R. B. Dodd. 1988. Electronic measurement of tree
canopy volume. Transactions of the ASAE 31(1): 264-272.
Hall A., D. W. Lamb, B. P. Holzapfel, J. P. Louis (2010) Within-season temporal
variation in correlations between vineyard canopy and winegrape composition and
yield Precision Agric (2011) 12:103–117
Hedley C. B., I. J. Yule, M. P. Tuohy, I. Vogeler (2009) Key Performance Indicators
For Simulated Variable Rate Irrigation of Variable Soils in Humid Regions Trans of
ASABE Vol. 52(5): 1575-1584
461
Heraud J.A., A, F. Lange (2009) Agricultural Automatic Vehicle Guidance from
Horses to GPS: How We Got Here, and Where We Are Going For presentation at the
2009 Agricultural Equipment Technology Conference Louisville, Kentucky, USA
Kitchen, N.R., Sudduth, K.A., Myers, D.B. Drummond, S.T., Hong, S.Y., (2005).
Delineating productivity zones on claypan soil fields apparent soil electrical
conductivity. Computers and Electronics in Agriculture, 46, 285-308.
Konopatzki, M.R.S., Souza, E.G., Nóbrega, L.H.P., Uribe-Opazo, M.A., Suszek, G.,
Rodrigues, S. and de Oliveira, E.F. 2009. Pear Tree Yield Mapping. Acta Hort.
(ISHS) 824:303-312 http://www.actahort.org/books/824/824_36.htm
Koch Br., R Khosla (2003) The Role of Precision Agriculture in Cropping Systems
Journal of Crop Production Vol. 9(1/2) 361-381
Kutter T., S. Tiemann, R. Siebert, S. Fountas (2011) The role of communication
and co-operation in the adoption of precision farming Prec. Agric (2011) 12:2–17
F3>=06$1'2)'*'7) )8"9/$1*'G)'7$�$1*'+)'H3/">31*'8)'I">J3'ABCCBE'Yield mapping of
cotton crop in Greece Paper presented in the HAICTA Conference, Athens.
Papageorgiou E.I. A.T.Markinos and T.A. Gemtos (2009) Soft Computing
Technique of Fuzzy Cognitive Maps to connect yield defining parameters with yield
in Cotton Crop Production in Central Greece as a basis for a decision support system
for precision agriculture application, Expert Syst. with Appl. 36(10)12399-413
Perry, C., S. Pocknee, O. Hansen, C. Kvien, G. Vellidis, and E. Hart. 2002.
Development and testing of a variable-rate pivot irrigation control system. ASAE
Paper No. 02-2290, ASAE, St. Joseph, MI.
Perry C.D. and A. W. Milton (2007) Variable-rate irrigation: concept to
commercialization Southern Conservation Agricultural Systems Conference June 25-
27,N.Florida Research & Education Center, Quincy.
Pozdnyakova L., D. Gime´nez, P. V. Oudemans (2005) Spatial Analysis of Cranberry
Yield at Three Scales Agron. J. 97:49–57
Qiao J,Akira S. S. Shibusawa, N. Kondo and E. Morimoto(2005) Mapping Yield
and Quality using the Mobile Fruit Grading Robot Biosystems Engng 90(2) 135-142
Schueller, J.K., J.D. Whitney, T.A. Wheaton, W.M. Miller, and A.E. Turner. (1999).
Low-cost automatic yield mapping in hand-harvested citrus. Computers and
Electronics in Agriculture 23:145-153.
Sethuramasamyraja B., S. Sachidhanantham1, R. Wample (2010) Geospatial
Modeling of Wine Grape Quality Indicators (Anthocyanin) for Development of
Differential Wine Grape Harvesting Technology International Journal of Geomatics
and Geosciences 1(3)
Tagarakis A., V. Liakos, S. Fountas, S. Koundouras, K. Aggelopoulou, T. Gemtos
(2010) Management zones delineation using fuzzy clustering techniques in
vines.EURAGENG conference, Clermont-Ferrand (France)
TAYLOR J.A., McBRATNEY A.B., WHELAN B.M., 2007. Establishing
management classes for broadacre agricultural production. Agron J 99(5) 1366-1376.
Tumbo S.D., M. Salyani, J. D. Whitney, T. A. Wheaton, W. M. Miller (2002)
Investigation of Laser and Ultrasonic Ranging Sensors for Measurements of Citrus
Canopy Volume Applied Engineering in Agriculture Vol. 18(3): 367–372
462