5782 research outputs found
Sort by
PHENO-DROP project ̶ advancing drought resilience in bread wheat through genomics and phenomics innovations
No full textBread wheat is a vital rainfed crop in both Italy and Serbia, where climate change has led to an increased frequency of droughts and extreme weather events, presenting serious threats to yield and yield stability. Traditional breeding approaches have achieved limited success in improving drought tolerance due to the complex, low-heritability nature of this trait and strong genotype-by-environment interactions. The project “New PHENO-ideotypes for DROught resilience in hexaPloid wheat – PHENO-DROP”, supported by the Ministry of Science, Technological Development and Innovation of the Republic of Serbia, under the Call for Joint Research and Innovation Projects 2024–2026 between Italy and Serbia, aims to address these challenges. The main focus of this bilateral project is to strengthen scientific excellence and innovation capacities through the exchange of practical and theoretical knowledge between research institutions in Italy and Serbia to improve hexaploid wheat breeding for drought conditions. PHENO-DROP integrates high-throughput phenomics, genomics, and bioinformatics to explore and validate drought-resilient pheno-ideotypes in a diverse panel of hexaploid wheat germplasm, including both landraces and modern bread wheat varieties. The project focuses on key traits related to Water Use Efficiency (WUE), such as root architecture, stomatal characteristics, osmotic adjustment, and canopy-level indices. It will assess genetic variability within wheat landraces to identify phenotypic traits, key genes, and regulatory mechanisms underlying water scarcity tolerance. Additionally, it will evaluate yield stability and identify associated phenotypic traits in a panel of modern Serbian and Italian wheat varieties across diverse environments. These outcomes will foster innovation in breeding approaches and support sustainable wheat production in drought-prone regions of the pan-Adriatic zone
Machine Learning (AutoML)-Driven Wheat Yield Prediction for European Varieties: Enhanced Accuracy Using Multispectral UAV Data
No full textAccurate and timely wheat yield prediction is valuable globally for enhancing agricultural
planning, optimizing resource use, and supporting trade strategies. Study addresses the
need for precision in yield estimation by applying machine-learning (ML) regression models
to high-resolution Unmanned Aerial Vehicle (UAV) multispectral (MS) and Red-Green-Blue
(RGB) imagery. Research analyzes five European wheat cultivars across 400 experimental
plots created by combining 20 nitrogen, phosphorus, and potassium (NPK) fertilizer treatments. Yield variations from 1.41 to 6.42 t/ha strengthen model robustness with diverse
data. The ML approach is automated using PyCaret, which optimized and evaluated
25 regression models based on 65 vegetation indices and yield data, resulting in 66 feature
variables across 400 observations. The dataset, split into training (70%) and testing sets
(30%), was used to predict yields at three growth stages: 9 May, 20 May, and 6 June 2022.
Key models achieved high accuracy, with the Support Vector Regression (SVR) model
reaching R2 = 0.95 on 9 May and R2 = 0.91 on 6 June, and the Multi-Layer Perceptron
(MLP) Regressor attaining R2 = 0.94 on 20 May. The findings underscore the effectiveness
of precisely measured MS indices and a rigorous experimental approach in achieving highaccuracy yield predictions. This study demonstrates how a precise experimental setup,
large-scale field data, and AutoML can harness UAV and machine learning’s potential
to enhance wheat yield predictions. The main limitations of this study lie in its focus on
experimental fields under specific conditions; future research could explore adaptability to
diverse environments and wheat varieties for broader applicability
Protein and oil content in NS soybean varieties at late sowing dates
No full textOptimalno vreme setve soje vezano je za temperaturu zemljišta. Soja se seje kada se temperatura zemljišta ustali na 10-12°C, a pomeranjem datuma setve i odabirom sorti različite dužine vegetacionog perioda može se uticati na ostvareni prinos zrna soje. Ranije sorte soje u ogledu ostvarile su najviše prinose pri setvi u aprilu mesecu, dok su kasnije sorte najviše prinose ostvarile pri majskoj setvi, pošto su ove sorte dočekale jesenje kiše zbog čega su ostvarile viši prinos. Najviši sadržaj proteina ostvaren je pri setvi 30. aprila, a najviši sadržaj ulja kod setve 10, maja. Cilj ovoga rada je sagledavanje prinosa, sadržaja proteina i ulja, kao i prinosa proteina i ulja po jedinici površine šest sorti soje pri kasnijim rokovima setve: 22. aprila, 30. aprila i 10. maja.The optimal time for sowing soybeans is related to soil temperature. Soybeans are sown when the soil temperature reaches 10-12°C, and by moving the sowing date and choosing varieties with different lengths of the growing season, the yield of soybeans can be influenced. The earlier soybean varieties in the experiment achieved the highest yields when sown in April, while the later varieties achieved the highest yields when sown in May, since these varieties received the autumn rains, which is why they achieved a higher yield. The highest protein content was achieved when sowing on April 30, and the highest oil content after sowing on May 10. The aim of this work is to analyze the yield, protein and oil content, as well as protein and oil yield per unit area of six soybean varieties at the later sowing dates: April 22, April 30 and May 10
EU Projects related to EMPHASIS
No full textThis is a material from the training delivered by the Plant Sciences (IBG-2) of Forschungszentrum Jülich GmbH, Germany on 12 March 2025 to the administration and research staff of the Institute of Field and Vegetable Crops, Serbia within the CROPINNO project. The material covers eligibility and feasibility for Infra-Serv Projects such as AgroServ, Microbes4Climate, PHENET, as well as information on key upcoming scientific events
Guidelines and best practices for research management and research funding
No full textGuidelines and best practices for research management and research funding is a deliverable of the CROPINNO project, funded as a HORIZON Coordination and Support Action by the European Commission under its Horizon Europe (HE) Programme. It is produced in the scope of Task 4.4 within Work Package 4: Strengthening research management and administration capacity. The obtained document will set a basis for future activities of Division for Project Management that IFVCNS has formally established in M13 of the CROPINNO project. Guidelines and best practices for research management and research funding was drafted by IFVCNS, which is the leader of T4.4, with input from all partners
Second report from workshops, Training Schools and STSMs
No full textSecond report from workshops, Training Schools and STSMs is a deliverable of the CROPINNO project, funded as a HORIZON Coordination and Support Action type by the European Commission under its Horizon Europe (HE) Programme. It is produced in the scope of Tasks 1.1. Scientific workshops, 1.2. Training Schools and T1.3. Short-Term Scientific Missions within Work Package 1: Strengthening scientific capacity. This document summarizes the activities completed within those three Tasks till the end of the duration of the project (M16-M36). The Report describes scientific workshops; training schools and short-term scientific missions organized within the second reporting period and in accordance with D6.1 Mobility Plan. The Report includes the following: Reports from workshops; Report from Training Schools (TS); Report from Short-Term Scientific Missions (STSM). The First report from workshops, Training Schools and STSMs was drafted by IFVCNS, which is the leader of the respective tasks, with input from all partners
Optimizing maize yield prediction using Gradient Boosting Machine with correlation-based feature selection
No full textAccurate yield prediction in maize is essential for improving decision-making in agronomic
management. This study presents a machine learning pipeline combining outlier detection,
correlation-based predictor filtering, and Gradient Boosting Machine (GBM) modeling. The
field experiment was conducted in 2024 at the Rimski Šanţevi research station (Novi Sad,Serbia) using a split-plot design. Four plant densities served as main plots, while six maize
hybrids were arranged as subplots. Yield and a set of spectral and canopy-related variables
were collected from the experiment, via combine harvester and UAV DJI Mavic 3E. The
dataset was split into training (70%) and testing (30%) subsets. Outliers in the training set
were assessed using the interquartile range (IQR) method, with no values exceeding the±3×IQR thresholds. Correlation analysis was applied to select variables with an absolute
Pearson correlation coefficient ≥ 0.40 with yield, resulting in the retention of nine predictors:Plant Height, Color Index of Vegetation Extraction, Excess Green minus Excess Red, Excess
Red Index, Mean Green-Red Vegetation Index, Normalized Green-Red Difference, Visible
Atmospherically Resistant Index, Visible Band Difference Vegetation Index and Percent
Vegetation Cover Area. A GBM was trained using repeated 10-fold cross-validation (5
repeats) with hyperparameter tuning. The best-performing model used 500 trees, a learning
rate of 0.01, interaction depth of 1, and a minimum of 5 observations per node. On thetraining set, the model achieved an RMSE of 613.3 and R² = 0.66. On the test set, RMSE was762.3 and R² = 0.744, indicating strong generalization. These results demonstrate that
vegetation indices derived solely from visible light bands — particularly those emphasizing
green-red contrast and canopy structure — can serve as effective predictors of maize yield.
The integration of basic filtering procedures with a robust ensemble model enabled accurate
prediction from RGB-based phenotyping data in a multi-factor field trial
Sugar profile of organic and conventional soybean seed
No full textSoja predstavlja jednu od najvažnijih proteinsko-uljarnih kultura, koja se koristi u ishrani ljudi i životinja, kao i u prehrambenoj i prerađivačkoj industriji. Cilj ovog istraživanja bio je da se analizira uticaj organskog i konvencionalnog sistema proizvodnje na profil šećera u semenu soje sorte Kaća, primenom metode HPLC-RI. Uzorci su prikupljeni na oglednom polju Instituta za ratarstvo i povrtarstvo u Novom Sadu tokom 2016. i 2017. godine. Dobijeni rezultati (sadržaj pentoza, heksoza, neredukujućih i redukujućih disaharida) izraženi su kao % ukupnih rastvorljivih šećera. U 2016. godini, konvencionalno proizvedeno seme značajno je premašilo organsko po ukupnom sadržaju rastvorljivih šećera.Soybean represents one of the most important protein – oilseed crops, used in human and animal nutrition, as well as in the food and processing industry. The aim of this study was to analyze the influence of organic and conventional production systems on the sugar profile in soybean seed of the Kaća variety, using the HPLC-RI method. Samples were collected from the experimental field of the Institute of Field and Vegetable Crops in Novi Sad during 2016 and 2017. The obtained results (content of pentoses, hexoses, non-reducing and reducing disaccharides) are expressed as a percentage of total soluble sugars. In 2016, conventionally produced seed significantly exceeded organic seed in total soluble sugar content
Analysis of the utilization of available quantities of tobacco stalks in Serbia
No full textPoljoprivredna proizvodnja u Srbiji raspolaže značajnim količinama biljne biomase koje se mogu iskoristiti u energetske svrhe. Jedan od potencijalnih izvora je biomasa stabljika duvana, naročito tipa Berlej. Cilj istraživanja bio je da se proceni raspoloživa količina stabljika ovog tipa i ispita njihov potencijal za energetsku valorizaciju. Analizirani su podaci Uprave za duvan o površinama pod različitim tipovima duvana – Berlej, Virdžinija i Orijental – s tim da su u proračune uključene samo stabljike tipa Berlej. Pri proračunu je korišćen prosek rasađivanja od 22.000 stabljika/ha i prosečna težina stabljike krupnolisnih duvana od 350 g.Agricultural production in Serbia generates significant amounts of plant biomass that can be used for energy purposes. One potential source is tobacco stalk biomass, particularly of the Burley type. The aim of this study was to estimate the available quantity of Burley stalks and examine their potential for energy valorization. Data from the Tobacco Administration on areas cultivated with different tobacco types – Burley, Virginia, and Oriental – were analyzed, with calculations including only Burley stalks. The calculation was based on an average planting density of 22,000 plants/ha and an average stalk weight of large-leaf tobaccos of 350 g
Plan for joint training course
No full textPlan for joint training course is a deliverable of the CROPINNO project, funded as a HORIZON Coordination and Support Action by the European Commission under its Horizon Europe (HE) Programme. It is produced in the scope of Task 1.4 within Work Package 1: Strengthening scientific capacity. This document describes a plan for joint training course on innovative approaches to climate-resilient crop improvement and production organized by IFVCNS and supported by CROPINNO partners. Plan for joint training course was drafted by IFVCNS, which is the leader of T1.4, with input from all partners