Produktivnost zemljišča je močno odvisna od globine tal. Ta podatek je nujno potreben pri izračunu količine hranil in vodo-zadrževalnih lastnosti tal. Za natančno oceno volumna tal, ki lahko zadrži hranila in vodo, potrebujemo tudi podatek o deležu skeleta v tleh. Ta dva podatka (globina in delež skeleta) sicer lahko pridobimo z uveljavljenimi metodami vzorčenja tal, a sta zaradi heterogenosti zemljišč pravilna le v točki vzorčenja. Za pridobitev bolj reprezentativnih podatkov bi morali na heterogenem zemljišču točkovno vzorčenje večkrat ponoviti, kar pa je drago, časovno zamudno in obenem tudi destruktivno. Alternativo predstavlja georadar, s katerim lahko hitro, neprekinjeno, neinvazivno in ponovljivo zaznavamo večje površine in znotraj njih opredelimo območja z različnimi talnimi značilnostmi. Princip georadarske metode temelji na zakonih prodiranja elektromagnetnih valov, ki jih v impulzih pošiljamo z oddajno anteno v tla. Ko elektromagnetno valovanje doseže mejo med različnimi materiali, se del vpadnega valovanja odbije in ga na površju zazna sprejemna antena. V članku smo poleg principov delovanja georadarja predstavili problematiko skeletnih tal in uveljavljenih metod vzorčenja. V pregledu literature smo navedli raziskave, s katerimi so avtorji uspešno določili globino (skeletnih) tal, prepoznali mejo med horizonti tal, napovedali delež skeleta in gline. V članku je predstavljen naš vidik, da se lahko georadar uspešno vključi v ustaljene metode dela in nam poleg bolj natančnih podatkov tudi prihrani nekaj časa. Vseeno je na tem področju, v primerjavi z uporabo georadarja na drugih področjih, še veliko neraziskanega.

Viticulture is considered to be the one of the most profitable agricultural sectors, with its main aim being the production of large quantities of high-quality grapes. Grapevine Flavescence dorée is a phytoplasma-associated disease that occurs in many grape-growing areas worldwide and is considered one of the most important grapevine diseases in Europe and is therefore subject to quarantine restrictions. A remote sensing early detection system could assist the experts and contribute to effective management of grapevine yellows foci, including limiting their spread. The potential of WorldView-2 satellite imaging for detecting grapevine yellows foci in a specific vineyard located in the coastal region of Slovenia was investigated. One vineyard with one of the most common red varieties in Slovenia (Refošk) was selected, and the health status of plants in three Trellis lines was determined. Each plant was assigned to one of four classes. The three classes with expressed symptoms were also pooled into a new class, for separate classifications. 39 spectral indices were calculated, yielding a total of 47 features, together with spectral bands. Sparse partial least squares discriminant analysis was used for feature extraction and determination of feature relevance. For both sets of classifications, 23 latent variables were extracted. Classifications using support vector machines achieved overall accuracies of above 75% on the test sets. A simple majority vote system was also implemented to reduce the influence of edge pixels and increase classification accuracy. Thus classification accuracies were increased to above 80% on the test set. The highest success rate in binary classification (87.7% on the test set) was attained through majority voting, as the majority of misclassifications were observed among the three classes of plants that exhibited expressed signs. These results showcase the applicability of satellite remote sensing for grapevine yellows detection, but more work is needed for such a system to become viable.

Commercial software is often prohibitively expensive, and therefore inaccessible for most users. Open-source software can facilitate analysis, but is usually less user friendly. The objective of this research was to develop, debug, and test SiaPy (Spectral imaging analysis for Python) software for image segmentation. It was evaluated on a hyperspectral dataset portraying one potato (Solanum tuberosum L.) cultivar with tolerance to drought. Pilot-case results confirm the efficacy of SiaPy utilization in the agricultural domain, especially in cases where multiple hyperspectral images need to be processed. Furthermore, it can be a great substitution for enterprise software, which are used to perform hyperspectral image segmentation.

Hyperspectral imaging is an effective tool for non-invasive plant disease detection and has been proven to work on various plant species. However, the technology is still not widely accepted and too expensive for use in certain applications. Less expensive multispectral cameras are available that can capture fewer spectral bands but may provide similar information in a particular domain. In this work, the focus is on the selection of the most prominent hyperspectral bands that could later be used for multispectral imaging applications. A solution based on convolutional neural networks and sparse sensing is proposed, and its performance is evaluated with classification metrics on a dataset consisting of healthy and diseased potato plants infected with Meloidogyne luci. The results indicated adequacy of the approach, as all tested classification algorithms achieved accuracy above 75%, even when only one spectral band was used.

The objective of our research was to test hyperspectral imaging as a method for early detection and discrimination of biotic and abiotic stress in maize. We investigated the individual and combined effects of wireworm feeding and drought stress on leaf spectral responses and on various morphological and physiological traits of maize plants, selecting two hybrids with different tolerance to drought. Physiological parameters were determined at three time points (14, 21 and 28 days after adding wireworms and changing watering regime), along with hyperspectral imaging. Most of the differences in physiological characteristics between treatments were detected on day 21, when drought was the main cause of the negative physiological outcome, while the presence of wireworms only caused lower relative chlorophyll content, resulting in lower combined stress damage in some treatments. The morphological data showed greater wireworm damage to hybrid ZP341 and a greater negative effect of combined stress on hybrid FuturiXX. Hyperspectral imaging detected pest infestation and drought stress before they were detected by classical methods, with the highest overall accuracy on day 14 (84.7%) and the lowest on day 28 (67%). It can therefore be used as a method for early detection of wireworm infestation and/or drought in maize.

Hyperspectral imaging is a popular tool used for non-invasive plant disease detection. Data acquired with it usually consist of many correlated features; hence most of the acquired information is redundant. Dimensionality reduction methods are used to transform the data sets from high-dimensional, to low-dimensional (in this study to one or a few features). We have chosen six dimensionality reduction methods (partial least squares, linear discriminant analysis, principal component analysis, RandomForest, ReliefF, and Extreme gradient boosting) and tested their efficacy on a hyperspectral data set of potato tubers. The extracted or selected features were pipelined to support vector machine classifier and evaluated. Tubers were divided into two groups, healthy and infested with Meloidogyne luci. The results show that all dimensionality reduction methods enabled successful identification of inoculated tubers. The best and most consistent results were obtained using linear discriminant analysis, with 100% accuracy in both potato tuber inside and outside images. Classification success was generally higher in the outside data set, than in the inside. Nevertheless, accuracy was in all cases above 0.6.

Root-knot nematodes (Meloidogyne spp.) are considered the most aggressive, damaging, and economically important group of plant-parasitic nematodes and represent a significant limiting factor for potato (Solanum tuberosum) production and tuber quality. Meloidogyne luci has previously been shown to be a potato pest having significant reproductive potential on the potato. In this study we showed that M. luci may develop a latent infestation without visible symptoms on the tubers. This latent infestation may pose a high risk for uncontrolled spread of the pest, especially via seed potato. We developed efficient detection methods to prevent uncontrolled spread of M. luci via infested potato tubers. Using hyperspectral imaging and a molecular approach to detection of nematode DNA with real-time PCR, it was possible to detect M. luci in both heavily infested potato tubers and tubers without visible symptoms. Detection of infested tubers with hyperspectral imaging achieved a 100% success rate, regardless of tuber preparation. The real-time PCR approach detected M. luci with high sensitivity.

Dry detention reservoirs are frequently used as a measure to alleviate flood risk. When implemented on agricultural land, they decrease the productive capacity of a rural landscape and negatively impact soil characteristics, crop quality and quantity, agricultural infrastructure (i.e., irrigation equipment), and agro-economics. The Slovenian government prepared a Detailed Plan of National Importance to ensure flood safety in the Lower Savinja Valley. The proposed plan foresees the implementation of a chain of 10 dry reservoirs (520 ha of agricultural land) to provide a higher level of flood protection for the cities of Celje and Las?ko. We explored how economic calculation can adequately address flood risk management initiatives and prevent adverse effects on high-quality agricultural land. The results of agro-economic scenarios for the study area showed that expected damage to agriculture production (crop yield) in a worst-case flood event occurring during the growing period would be EUR 1.7 million. The most exposed farms would suffer damage in the range of three to five average annual salaries. These agro-economic findings indicate the need for discussion on agriculture-focused mitigation measures before the construction of flood protection measures.

Root-knot nematodes are considered the most important group of plant-parasitic nematodes due to their wide range of plant hosts and subsequent role in yield losses in agricultural production systems. Chemical nematicides are the primary control method, but ecotoxicity issues with some compounds has led to their phasing-out and consequential development of new control strategies, including biological control. We evaluated the nematicidal activity of Bacillus firmus I-1582 in pot and microplot experiments against Meloidogyne luci. I-1582 reduced nematode counts by 51% and 53% compared to the untreated control in pot and microplot experiments, respectively. I-1582 presence in the rhizosphere had concurrent nematicidal and plant growth-promoting effects, measured using plant morphology, relative chlorophyll content, elemental composition and hyperspectral imaging. Hyperspectral imaging in the 400–2500 nm spectral range and supervised classification using partial least squares support vector machines successfully differentiated B. firmus-treated and untreated plants, with 97.4% and 96.3% accuracy in pot and microplot experiments, respectively. Visible and shortwave infrared spectral regions associated with chlorophyll, N–H and C–N stretches in proteins were most relevant for treatment discrimination. This study shows the ability of hyperspectral imaging to rapidly assess the success of biological measures for pest control.

Crop infestation with root-knot nematodes (RKN) and water deficiency lead to similar visible symptoms in the plant canopy. Identification of biotic or abiotic stress origin is therefore a problem, and currently the only reliable methods for determination of RKN infestation are invasive and applicable only for point-searches. In this study the applicability of hyperspectral remote sensing for early identification of drought stress and RKN infestations in tomato plants was tested. A four-stage image and data management pipeline was established: (1) image acquisition, (2) data extraction, (3) pre-processing, and (4) processing. •This pipeline reduces atmospheric impacts, facilitates data extraction (by using specially designed spectral libraries and supervised classification procedures), diminishes the impact of viewing geometry, and emphasized small spectral variations not apparent in the raw data.•By combining partial least squares - discriminant analysis and support vector machines with time series analysis, we achieved up to 100% classification success when determining watering regime and infestation, and their severity.•This pipeline could be at least partially automated, thus facilitating high throughput identification of stress origin in plants. Furthermore, the same pipeline could be applied to hyperspectral phenotyping procedures, which are gaining importance in breeding programs.

ABSTRACTTraditional agricultural plant pest and disease management practices are based on visible characteristics and require that plants are checked individually, making these practices time consuming and therefore costly. Plant pests and diseases also often exhibit a heterogeneous distribution, making detection more difficult. Remote sensing methods enable comparatively accurate detection of pests and diseases over larger areas. Furthermore, because remote sensing sensors utilize light outside the human visible spectrum, presymptomatic detection becomes possible, thus facilitating timely, appropriate and spatially accurate management practices. Because remote sensing systems generate large amount of data, novel data analysis methods, such as machine learning, were introduced to plant protection. While pest and disease detection is possible using individual sensors, best results can be obtained by combining different sensors, utilizing different spectral ranges or physiological responses to light. A large amount of data and information has been generated in the past, but this research has mostly been focused on individual pathogens. Future research will have to focus on combined infections or infestations, and include abiotic stressors as well. Key words: Remote sensing, plant protection, hyperspectral, multispectral, thermal, fluorescence, precision agriculture IZVLEČEKVelikokrat tradicionalni pristopi varstva rastlin pred rastlinskimi boleznimi in škodljivci temeljijo na vidnih simptomih, ki vključuje redno pregledovanje posameznih rastlin. Postopki so zato lahko dolgotrajni in s tem dragi. Bolezni in škodljivci imajo v prostoru pogosto heterogeno razporeditev, kar otežuje njihovo odkrivanje. Metode daljinskega zaznavanja omogočajo razmeroma natančno odkrivanje škodljivcev in bolezni na večjih območjih. Ker uporabljajo senzorji daljinskega zaznavanja tudi svetlobo izven nam vidnega spektra, je možno tudi zgodnje odkrivanje, t.j. odkrivanje pred razvojem vidnih znakov bolezni. To omogoča pravočasno, ustrezno in prostorsko natančno upravljanje z boleznimi in škodljivci. Sistemi daljinskega zaznavanja ustvarjajo velike količine podatkov, zato so bile v varstvo rastlin uvedene sodobne metode za analizo podatkov, na primer strojno učenje. Čeprav je možno zaznava bolezni in škodljivcev z uporabo posameznih senzorjev, lahko dosežemo najboljše rezultate z združevanjem različnih senzorjev, torej z uporabo različnih spektralnih območij ali fizioloških odzivov na svetlobo. Dosedanje raziskave so bile osredotočene na posamezne škodljivce in bolezni. Prihodnje raziskave se bodo morale osredotočiti na kombinirane okužbe ter vključevati tudi abiotske stresorje. Ključne besede: daljinsko zaznavanje, varstvo rastlin, hiperspekter, multispekter, toplotno slikanje, fluorescenca, precizno kmetijstvo

In karst landscapes stony soils have little water holding capacity; the rational use of water for irrigation therefore plays an important management role. Because the water holding capacity is not homogenous, precision agriculture approaches would enable better management decisions. This research was carried out in an experimental vineyard grown in an artificially transformed karst terrain in Dalmatia, Croatia. The experimental design included four water treatments in three replicates: (1) fully irrigated, based on 100% crop evapotranspiration (ETc) application (N100); (2 and (3) deficit irrigation, based on 75% and 50% ETc applications (N75 and N50, respectively); and (4) non-irrigated (N0). Hyperspectral images of grapevines were taken in the summer of 2016 using two spectral-radiance (W sr−1 m−2) calibrated cameras, covering wavelengths from 409 to 988 nm and 950 to 2509 nm. The four treatments were grouped into a new set consisting of: (1) drought (N0); and (2) irrigated (the remaining three treatments: N100, N75, and N50). The images were analyzed using Partial Least Squares-Discriminant Analysis (PLS-DA), and treatments were classified using PLS-Single Vector Machines (PLS-SVM). PLS-SVM demonstrated the capability to determine levels of grapevine drought or irrigated treatments with an accuracy of more than 97%. PLS-DA identified relevant wavelengths, which were linked to O–H, C–H, and N–H stretches in water, carbohydrates and proteins. The study presents the applicability of hyperspectral imaging for drought stress assessment in grapevines, even though temporal variability needs to be taken into account for early detection.

The aim of this study was to increase the insight on the application of hyperspectral imaging as a non-destructive method in the prediction of the principal parameters that compose technological and phenolic maturity (i.e. sugar concentration, total acidity, total phenols, and anthocyanin content) in wine grapes. The research was conducted in a Babic (Croatian autochthonous cultivar) vineyard grown in artificially transformed karst terrain (Croatia). The hyperspectral images were recorded by a hyperspectral imaging system Hyspex VNIR-1600 and Swir-384 (Norsk Elektro Optikk, Norway) with a spectral range from 400-1000 nm (VNIR), and 1000-2500 nm (SWIR). PLS regression enables us to predict the amount of sugar and acid in grapes with acceptable accuracy $({\mathrm {R_{sugar}}}^{2} = 0.92$, and ${\mathrm {R_{acid}}}^{2} = 0.83)$. By using hyperspectral imaging measurements of sugar content in grapes can become non-invasive, and by using mobile remote sensing systems, cover entire vineyards.

The work aims to investigate the effects of different soil management strategies on carbon sequestration and total nitrogen in areas of vineyards suffering from loss of soil functionality. Treatments, selected for inter-row management, to re-install soil functionality were based on compost or other organic amendments (COMP), green manure (GM), and dry mulching (DM) strategies using winter legumes and cereals. Cover crops were seeded in fall and mown in late spring, leaved in the ground for mulching in DM or incorporated into the uppermost soil layers in GM. Such approaches were investigated in six vineyards in Italy, six in France, and two vineyards in Slovenia and Turkey. The results showed that COMP significantly increased total organic carbon (TOC) and total nitrogen (Ntot) in the topsoil after one year of application. Also DM tends to increase significantly TOC in the topsoil, but only after two years. Modelling 20-year carbon stock dynamics in Italy vineyards, the average increase resulted 0.49, 0.34, 0.21 and 0.03 Mg C ha-1 yr-1 for COMP, DM, GM and control, respectively.

Crop plants are subjected to various biotic and abiotic stresses. Both root-knot nematodes (biotic stress) and water deficiency (abiotic stress) lead to similar drought symptoms in the plant canopy. In this work, hyper-spectral imaging was used for early detection of nematode infestation and water deficiency (drought) stress in tomato plants. Hyperspectral data in the range from 400 to 2500 nm of plants subjected to different watering regimes and nematode infestation levels were analysed by partial least squares- discriminant analysis (PLS-DA) and partial least squares - support vector machine (PLS-SVM) classification. PLS-SVM classification achieved up to 100% accuracy differentiating between well-watered and water-deficient plants, and between 90 and 100% when identifying nematode-infested plants. Grouping the data according to the time of imaging increased the accuracy of classification. Shortwave infrared spectral regions associated with the O-H and C-H stretches were most relevant for the identification of nematode infested plants and severity of infestation. This study demonstrates the capability of hyperspectral imaging to identify and discriminate between biotic and abiotic plant stresses.

This multidisciplinary research work evaluated the effects of soil erosion on grape yield and quality and on different soil functions, namely water and nutrient supply, carbon sequestration, organic matter recycling, and soil biodiversity, with the aim to understand the causes of soil malfunctioning and work out a proper strategy of soil remediation. Degraded areas in nineteen organically farmed European and Turkish vineyards resulted in producing significantly lower amounts of grapes and excessive concentrations of sugar. Plants suffered from decreased water nutrition, due to shallower rooting depth, compaction, and reduced available water capacity, lower chemical fertility, as total nitrogen and cation exchange capacity, and higher concentration of carbonates. Carbon storage and organic matter recycling were also depressed. The general trend of soil enzyme activity mainly followed organic matter stock. Specific enzymatic activities suggested that in degraded soils, alongside a general slowdown in organic matter cycling, there was a greater reduction in decomposition capacity of the most recalcitrant forms. The abundance of Acari Oribatida and Collembola resulted the most sensitive indicator of soil degradation among the considered microarthropods. No clear difference in overall microbial richness and evenness were observed. All indices were relatively high and indicative of rich occurrence of many and rare microbial species. Dice cluster analyses indicated slight qualitative differences in Eubacterial and fungal community compositions in rhizosphere soil and roots in degraded soils. This multidisciplinary study indicates that the loss of soil fertility caused by excessive earth movement before planting, or accelerated erosion, mainly affects water nutrition and chemical fertility. Biological soil fertility is also reduced, in particular the ability of biota to decompose organic matter, while biodiversity is less affected, probably because of the organic management. Therefore, the restoration of the eroded soils requires site-specific and intensive treatments, including accurately chosen organic matrices for fertilization, privileging the most easily decomposable. Restoring soil fertility in depth, however, remain an open question, which needs further investigation.

This research was carried out in an experimental vineyard grown in artificially transformed karst terrain (Croatia). The experimental design included four water treatments in three replicates: 1) fully watered or based on 100% evapotranspiration (ETc) application; 2) regulated deficit irrigation based on 75% and 50% ETc applications; and 3) non-watered. Hyperspectral images of grapevines were taken in the summer of 2016 using two spectral-radiance (W·sr−1·m−2) calibrated cameras, covering wavelengths from 409 to 988 nm and 950 to 2509 nm. The four treatments were grouped into two new sets: 1) drought (NO); and 2) watered (the remaining three treatments). The watered group was then split into two new sets: 1) 50% treatment; and 2) 75+100% treatment. The images were analyzed using Partial least squares – discriminant analysis (PLS-DA), and Single Vector Machines (SVM). The PLS-DA demonstrated the capability to determine levels of grapevine drought or watered groups with 70 - 80% accuracy. A similar success rate was achieved in distinguishing the 50% group from the 75+100% group.

Near-infrared (NIR) hyperspectral imaging was explored as a rapid and non-destructive method of investigating seed quality parameters such as seed viability and variation in tomato seed lots. The seed lots differed with year of production and variety. Four tomato varieties: Cal J, Monprecus, NCL and Chiuri from 2013, 2014 and 2015 were used in the study. The extracted NIR hyperspectral data from 975 to 2500 nm were analysed by principal component analysis (PCA) and partial least squares- discriminant analysis (PLS-DA). No distinct patterns of separation between viable and non-viable tomato seeds were revealed by the PCA. Our findings showed a pattern of separation in the tomato seed lots due to production years and varieties. The PLS-DA showed the ability to predict with 100 percent accuracy for varietal class membership when only the seeds of a single harvest year were included in the model. The accuracy from PLS-DA on pooled samples (all seeds from all varieties) predicted varietal class membership in the range from 34 to 88 percent. High variation in the seed lots could have caused high variation in the predicted varietal class membership. The NIR regions with chemical information from C-H, N-H and O-H had influence on the PCA and PLS-DA models. The study presents the prospects of using NIR hyperspectral imaging in varietal identification studies of tomato seeds though we recommend a thorough validation of models. (C) 2016 Elsevier B.V. All rights reserved.