| Sažetak | Digitalizacija vinogradarske proizvodnje u posljednjih tridesetak godina doprinosi brzom
razvoju novih tehnologija za praćenje proizvodnje, ali i prihvaćanju i usvajanju novih
spoznaja vezanih uz varijabilnost vinograda i mogućnost iskorištavanja varijabilnosti za
povećanje ekonomske učinkovitosti proizvodnje. Korištenjem bespilotnih letjelica
opremljenih multispektralnim kamerama te analizom podataka putem vegetacijskih
indeksa moguće je donositi kvalitetnije odluke za upravljanje vinogradarskom
proizvodnjom, ali i maksimizirati učinke proizvodnje u agronomskom, enološkom i
ekonomskom smislu. Cilj ovog istraživanja bio je utvrditi povezanost vegetacijskih indeksa
(NDVI, NDRE, OSAVI) sa svojstvima vinove loze i grožđa kako bi se učinkovito
identificirale dvije različite kvalitativne zone grožđa unutar istog vinograda, s ciljem
selektivne berbe grožđa i povećanja prihoda od vinogradarske i vinarske proizvodnje.
Istraživanje je provedeno tijekom 2019. i 2020. godine na području Zagrebačke županije
na četiri različite lokacije i četiri različite površine vinograda manje od 1 ha (0,33 ha; 0,47
ha; 0,65 ha i 0,93 ha). U istraživanje su bile uključene dvije sorte vinove loze- 'Kraljevina'
(Vitis vinifera L.) u vinogorju Zelina i 'Pinot crni' (Vitis vinifera L.) u vinogorju Plešivica-
Okić. Podaci su prikupljani bespilotnom letjelicom opremljenom multispektralnom
kamerom u tri različite fenofaze razvoja vinove loze te obrađeni i analizirani u ArcGIS
programskom paketu. Provedeno je i ručno uzorkovanje i mjerenje sastavnica prinosa i
kvalitativnih svojstava grožđa na unaprijed odabranim ciljanim trsovima. Podaci su
analizirani postupcima deskriptivne, inferencijalne te multivarijatne statistike, kako bi se
dobio najprediktivniji vegetacijski indeks za kvalitativno zoniranje vinograda na svakoj
lokaciji i u obje godine. Metodom ankete i desk istraživanja prikupljeni su podaci za
analizu ekonomske opravdanosti ulaganja u kvalitativno zoniranje i selektivnu berbu te su
izrađene financijske analize investicije za dva moguća scenarija- samostalna i uslužna
provedba kvalitativnog zoniranja.
Rezultati istraživanja mogućnosti primjene vegetacijskih indeksa za kvalitativno zoniranje
vinograda pokazali su kako su vegetacijski indeksi učinkovit alat za procjenu bujnosti i
varijabilnosti u vinogradu te kako uspješno opisuju i povezuju sastavnice prinosa i
kvalitativna svojstva grožđa s rezultatima spektralnih mjerenja. NDVI kao najviše korišten
vegetacijski indeks u dosadašnjim istraživanjima i u se ovom istraživanju pokazao
najprediktivnijim vegetacijskim indeksom za određivanje zona bujnosti i kvalitativno
zoniranje vinograda. NDRE, iako vrlo skromno korišten u dosadašnjim istraživanjima za
određivanje zona bujnosti i kvalitativno zoniranje vinograda u ovom istraživanju pokazao
se dovoljno prediktivnim za praktične svrhe provedbe selektivne berbe u vegetacijskoj
sezoni u kojoj se prikupljaju podaci. S obzirom na termin prikupljanja podataka za
kvalitativno zoniranje putem vegetacijskih indeksa NDVI i NDRE, potrebno je višegodišnje
praćenje određenog vinograda i analiza mogućih interakcija u samom vinogradu kako bi
se definirao termin prikupljana podataka za učinkovito kvalitativno zoniranje i selektivnu
berbu. OSAVI, kao najmanje korišten vegetacijski indeks u dosadašnjim istraživanjima i u
ovom se istraživanju pokazao najmanje prediktivan, odnosno niti u jednom terminu
prikupljanja podataka kao niti na jednoj lokaciji provedbe istraživanja, OSAVI nije bio
najprediktivniji vegetacijski indeks za određivanje zona bujnosti i kvalitativno zoniranje
vinograda. Primjena vegetacijskih indeksa kao osnove za kvalitativno zoniranje vinograda
i selektivnu berbu grožđa u vinogradima do 1ha, s ciljem proizvodnje dva različita tipa
„super premium“ vina od iste sorte grožđa i iz istog vinograda, može dovesti do povećanja
financijske koristi za vinogradara. Dodatno, utvrđeno je da se korištenje usluge
specijalizirane tvrtke za snimanje bespilotnom letjelicom i obradu i analizu prikupljenih
podataka pokazuje financijski isplativijim rješenjem od ulaganja u opremu i samostalnog
provođenja kvalitativnog zoniranja od strane vinogradara. |
| Sažetak (engleski) | The digitization of viticulture in the last thirty years contributes to the rapid development of
new technologies for monitoring production and the acceptance and adoption of new
knowledge about the variability of vineyards and the possibility of using variability to
increase the economic efficiency of production. The uniform vineyard management can be
justified by the simplicity of carrying out all vineyard operations as long as the winegrower
is satisfied with the final grape quality. Identifying different quality zones based on
variations within the vineyard can allow winegrowers to obtain higher revenues from grape
growing and wine production. By using unmanned aerial vehicles (UAV) equipped with
multispectral cameras and analysing the collected data using vegetation indices, it is
possible to make better decisions for differentiated vineyard management, maximizing the
agronomic, oenological and economic impact of viticulture production. The main objective
of this study was to determine the most predictive vegetation index for grape quality
zoning, among three different vegetation indices (NDVI, NDRE, and OSAVI) at three
different grapevine growth stages, which can be efficiently used in commercial vineyards
for selective harvesting and the production of different wine types. In addition, an
economic analysis of the costs and revenues of implementing grape quality zoning and
selective harvesting in small vineyards (up to 1 ha) was performed, along with the
calculations of the potential revenue increases after selective harvesting and production of
different wine types from the same grapevine variety and vineyard.
The first chapter, Introduction, focuses on the problem description and presents the main
aspects of quality zonal management and the main hypotheses and objectives of the
study. The hypotheses that were established and tested in this study are: (1) there is a
difference in the relationship between vegetative indices (NDVI, NDRE, OSAVI) and
vigour at three different grapevine growth stages (GS) for two grapevine varieties; (2)
there is a relationship between different vigour zones and vegetative, yield, and grape
quality components; (3) there is economic justification for using vegetative indices for
vineyard quality zonal management and selective harvesting in vineyards up to 1 ha.
The second chapter Overview of the Previous Research focuses on related previous
research on vineyard variability, precision viticulture, remote sensing, vegetation indices,
vineyard quality zonal management and relevant studies on the economic efficiency of
precision viticulture. The main reason for studying vineyard variability is its influence on
grape yield and quality. Variability can be caused by various factors (natural factors,
human actions, factors beyond control, etc.), but regardless of its source, monitoring and
managing of variability has the goal of helping winegrowers to achieve the greatest
possible efficiency in viticulture production. In order to monitor the different parameters
and the related changes in the vineyard throughout the growing season, many new tools
and technologies are being developed and used in viticulture. There are two main groups
of technologies that are being used: data-intensive technologies and automated
technologies, which are differentiated according to the main outcomes that result from the
use of each technology. Data-intensive technologies, such as vigour and yield monitoring
and mapping, soil condition mapping, and environmental data collection can improve
understanding of the factors causing variability, but also require additional skills to use the
technologies effectively and it is much more difficult to estimate economic efficiency. Much
research has been done on the capabilities of UAVs. UAVs have been used to assess
grape yield and quality, water stress, photosynthetic activity, disease, pest and weed
incidence, but also to detect missing vines, estimate vine height, estimate damage caused
by flooding and drought, etc. In the last twenty years, much research has been done with
vegetation indices (mainly NDVI), and these spectral vegetation measurements have
been used to describe vine and grape characteristics such as vigour, yield, grape quality,
health status, etc. Previous research has shown that vegetation indices can be good
predictive tools for vigour-based quality management zones and for assessing vineyard
variability. Some of the most commonly used vegetation indices in viticulture are NDVI,
NDRE, and OSAVI. The relationship between vigour zones and yield and grape quality
components has been confirmed in many previous studies, which showed that high vigour
zones delineated mainly with NDVI had higher yield per vine, lower sugar concentration,
and higher total titratable acidity than vines from low or medium vigour zones. Selective
harvesting is defined as the split-picking of grapes at harvest according to different
yield/quality criteria with the aim of producing different products to take advantage of the
observed variability in vineyard performance. In the economic analysis, the investment in
the new technology must be considered in relation to the expected benefits for the
winegrowers. The potential profitability of precision farming can only be evaluated in
comparison to management without precision farming. The economic aspect of selective
harvesting mainly refers to the differentiation of grapes from different quality zones with
different future wine prices.
The third chapter Materials and Methods focuses on all the information on how the
research was conducted. The research was conducted in 2019 and 2020 in the area of
Zagreb County in four different locations and on four different vineyard areas of less than
1 ha (0.33 ha, 0.47 ha, 0.65 ha and 0.93 ha). Two grapevine varieties were included in the
study - 'Kraljevina' (Vitis vinifera L.) in the Zelina winegrowing region and 'Pinot noir' (Vitis
vinifera L.) in the Plešivica winegrowing region. Using a UAV DJI Inspire 1 Pro equipped
with a multispectral camera (Micasense RedEdge.MXTM, Seattle, WA, USA) multispectral
images were acquired each year at the experimental sites at three different grapevine GS:
berries beginning to soften, Brix starts increasing (GS 34); berries with intermediate Brix
values (GS 36); and berries harvest-ripe (GS 38). After each flight, images were
processed using ATLAS MicaSense app (MicaSense, Seattle, DC, USA), Pix4D software
package (Pix4D, Lausanne, Switzerland) and ArcGIS software package (Environmental
Systems Research Institute (ESRI) (2012), ArcGIS Release 10.1, Redlands, CA, USA).
Interpolated maps representing two vigour zones using NDVI, NDRE, or OSAVI were
generated for each site and flight and used for subsequent statistical analyses. Manual
sampling and measurements of yield and grape quality components were performed on
pre-selected target vines. The number of target vines was determined according to the
area of vineyard. Each target vine was assigned to a vigour zone (high or low) based on
the two zones previously defined. In addition, for each map created, the percentage of
vineyard area in low and high vigour zones was calculated to allow for economic efficiency
calculations for grape quality zoning and selective harvesting based on the vegetation
indices. The collected grape quality components were subjected to a clustering procedure
at each site and in each year, with the K-means clustering procedure set for two clusters.
In this way, two different grape quality clusters (clusters of better and inferior grape
quality) were obtained and used for comparison with the target vine vigour zone structures
(categorical variable) based on NDVI, NDRE, and OSAVI vigour maps at three different
grapevine GSs. The target vines belonging to the low vigour zone (green in vigour maps)
were characterised with better grape quality components, while the target vines belonging
to the high vigour zone (red in vigour maps) were characterised with inferior grape quality
components. The vegetative index whose classification structure of target vines vigour
(categorical variable) most closely matched the cluster structure of determined grape
quality was considered the most predictive. Using the survey method and secondary desk
research, data were collected for the analysis of the economic efficiency of investment in
quality zonal management and selective harvesting. Financial analyses of the investment
for two scenarios: the winegrower performing quality zonal management himself and the
winegrower as a user of a commercial service for quality zonal management.
In the chapters Results and Discussion, all the research results are presented and
discussed: interpolated maps for each site, each vegetation index, each GS and both
years; results of the clustering procedure and of the overlap between two classification
structures (vigour and grape quality) for the most predictive vegetation index, together
with the descriptive statistical analysis and the results of testing the statistical significance
of the differences between the two vigour zones based on the most predictive vegetation
index. Finally, the results of the economic efficiency and financial analysis are presented.
All costs (fixed and variable) and potential revenues from grape and wine production after
selective harvesting were calculated. Investment analysis were prepared for all sites and
two possible scenarios, profit or loss statements and relevant financial criteria for
investment analysis were calculated (NPV, IRR, payback period, discounted payback
period, mIRR). A two-year study on the possibility of using three different vegetation
indices- NDVI, NDRE and OSAVI- for grape quality zoning has shown that they are
effective tools for assessing vigour and variability in small vineyards in Zagreb County and
that they can successfully describe yield and grape quality components and link them to
the results of spectral measurements. However, there are differences in the effectiveness
of estimating the relationship between vigour, yield, and grape quality components
depending on the vegetation index used and the grapevine’s GS. The NDVI as the most
commonly used vegetation index in previous studies, in this study also proved to be the
most predictive for delineating vigour zones and grape quality zoning of vineyards. NDVI
was the most predictive vegetation index four times (out of seven) and at all sites. In
addition, the percentage of overlap of classification quality structures was 81%-93%,
making it the most predictive vegetation index for quality zoning. Further research is
needed to determine the ideal time period for UAV data acquisition depending on climatic
conditions during the growing season. Although NDRE has been used very modestly in
previous studies for delineating vigour zones and grape quality zoning of vineyards, it
proved to be the most predictive vegetation index three times (out of seven) at three sites.
At these sites, NDRE successfully delineated grape quality zones at later grapevine GSs,
and the zones were statistically significant. It can be said that the use of NDRE for grape
quality zoning of vineyards is sufficiently predictive for purposes such as selective
harvesting in the growing season in which the data are collected. OSAVI, the least used
vegetation index in previous studies, was also found to be the least predictive in this study
for delineating vigour zones and grape quality zoning of vineyards. OSAVI was not the
most predictive at any grapevine GS or any study site. These results are consistent with
previous studies in which OSAVI was used primarily to estimate the nitrogen content in
grapevines for targeted nitrogen fertiliser applications. The potential economic efficiency
of grape quality zoning and selective harvesting was evident in this study. The potential
sales revenue from the production of “super-premium” wines after selective harvest was
14% to 45% higher than the total revenue that could be generated from the production of
“quality” wines at all sites. Considering that all the wineries are small, family-owned
boutique wineries and the selling prices of their wines were above average, the important
finding is that they can use high- and low-quality zones to produce two types of “superpremium”
wines. In terms of investment, both grape quality zoning scenarios (A- the
winegrower performs grape quality zoning himself, B- the winegrower uses a commercial
service for grape quality zoning) resulted in potentially higher profits and the financial
criteria suggested that this would be an investment that can bring higher profits in a very
short payback period, despite the higher costs required to implement grape quality zoning
and selective harvesting. Using a commercial service provider for grape quality zoning
resulted in even higher potential profits and lower risks for the winegrower.
It can be concluded that vegetation indices NDVI and NDRE are effective tools for
vineyard quality zonal management and can enhance the economic efficiency of
viticulture through selective harvesting and the production of two types of “super-premium”
wines from small vineyard areas up to 1 ha. Further research could be considered for the
use of OSAVI on images where row segmentation has already been done and where
grapevine vegetation and soil data are already separated. In addition, the ideal UAV data
acquisition period for grape quality zoning still needs to be studied. |
