PGA PROGRAMA DE PÓS-GRADUAÇÃO EM AGRONOMIA CENTRO DE CIENCIAS AGRARIAS Teléfono/Ramal: No informado
Disertación/Tesis

Clique aqui para acessar os arquivos diretamente da Biblioteca Digital de Teses e Dissertações da UEM

2026
Disertaciones
1
  • NICOLE GHINZELLI VEDANA
  • -

  • Líder : MARCOS RAFAEL NANNI
  • MIEMBROS DE LA BANCA :
  • AMANDA SILVEIRA REIS
  • LUÍS GUILHERME TEIXEIRA CRUSIOL
  • MARCOS RAFAEL NANNI
  • Data: 26-feb-2026


  • Resumen Espectáculo
  • -

2
  • RODRIGO ORSINI POLA
  • Crescimento, desenvolvimento e controle químico de feijão-de-rôla (Macroptilium lathyroides)

  • Líder : RUBEM SILVERIO DE OLIVEIRA JUNIOR
  • MIEMBROS DE LA BANCA :
  • DENIS FERNANDO BIFFE
  • JAMIL CONSTANTIN
  • PAULO VINICIUS DA SILVA
  • RUBEM SILVERIO DE OLIVEIRA JUNIOR
  • Data: 27-feb-2026
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The increasing occurrence of Macroptilium lathyroides (L.) Urb. (phasey bean) in Brazilian
    agrosystems, particularly in soybean crops, imposes new challenges to weed management due to the
    scarcity of bioecological information and the lack of herbicides registered for the species. This
    study aimed to investigate the morphophysiology and chemical control strategies of this weed.
    Greenhouse experiments were conducted and structured into two chapters. The first chapter
    characterized the growth, phenology, and reproductive potential of the species. Results
    demonstrated that M. lathyroides exhibits an indeterminate growth habit, maintaining its
    development even after seed production, with rapid initial establishment, reaching flowering at 54
    days after emergence (DAE). Growth analysis revealed a strategic biomass allocation, prioritizing
    accumulation in stems (50.7% of total dry matter) from 33 DAE onwards, which supports its hardiness
    and regrowth capacity. The high reproductive potential, estimated at 2,938 seeds per plant, defines
    a critical management period prior to 50 DAE. The second chapter evaluated herbicide efficacy in
    pre- and post-emergence. In pre-emergence, Protoporphyrinogen Oxidase (PROTOX) inhibitors,
    specifically oxyfluorfen and flumioxazin, completely prevented species establishment. In early
    post-emergence, response variability was observed among populations, highlighting differential
    tolerance to bentazon. Synthetic auxins and rapid-contact PROTOX inhibitors (saflufenacil,
    tiafenacil, and epyrifenacil) were effective, while glyphosate required high doses for satisfactory
    control. For adult plants, the insufficiency of glyphosate alone was confirmed, requiring
    association with robust broadleaf herbicides, such as synthetic auxins, to prevent regrowth. The
    herbicide epyrifenacil stood out for its excellent efficacy when applied alone even at advanced
    stages. It is concluded that the management of
    M. lathyroides requires the integration of preventive tactics and the correct positioning of
    herbicides with alternative mechanisms of action, aiming to mitigate its high competitive and
    reproductive capacity.

3
  • DAIANE DE FATIMA DA SILVA HAUBERT
  • Classificação de solos e predição da granulometria e do carbono orgânico usando dados hiperespectrais e modelos de Machine Learning

  • Líder : MARCOS RAFAEL NANNI
  • MIEMBROS DE LA BANCA :
  • LUÍS GUILHERME TEIXEIRA CRUSIOL
  • MARCOS RAFAEL NANNI
  • RENAN FALCIONI
  • Data: 24-jun-2026
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Soil classification and the prediction of their physical and chemical attributes remain a challenge when relying solely on field descriptions and conventional laboratory analyses, particularly at high sampling densities. For this reason, there is growing interest in complementary tools to assist in this task, especially through spectroscopy combined with machine learning (ML) algorithms. Thus, this study aims to investigate the potential of spectral reflectance from intact soil profiles, combined with ML models, for classifying soil classes and pedogenic horizons, as well as for predicting attributes. Six monoliths were collected in the Central-South and Northwest regions of the state of Paraná and classified according to the Brazilian Soil Classification System (SiBCS). For each monolith, 800 spectra were acquired using a non-imaging VIS-NIR-SWIR spectroradiometer (350 - 2,500 nm), and 800 spectra were extracted from SWIR hyperspectral images (1,200 - 2,450 nm). For the classification of soil classes, the data were divided into two subsets: reflectance of surface horizons and reflectance of subsurface horizons. Spectral reflectance was analyzed using principal component analysis (PCA) and processed using the following algorithms: K-nearest neighbors (KNN), Random Forest (RF), Decision Tree (DT), Neural Network (NN), and Gradient Boosting (GB), with cross-validation. Performance was evaluated using the following parameters: Accuracy (CA), F-Score, Area Under the Curve (AUC), and Matthews’ Correlation Coefficient (MCC). To predict the attributes (particle size distribution and soil organic carbon - SOC), the reflectance samples were correlated with data obtained in the laboratory and analyzed using the following ML models: KNN, RF, GB, and Partial Least Squares Regression (PLSR). Prediction performance was evaluated using the following parameters: Root Mean Square Error (RMSE), Performance-to-Deviation Ratio (RPD), systematic error (BIAS), and coefficient of determination (R²). PCA indicated that most of the spectral variance was captured by a single dominant axis, with a clearer separation between suborders in the SWIR space than across the entire VIS-NIR-SWIR band. For soil class classification, subsurface spectra performed better than surface spectra, and SWIR performed better than VIS-NIR-SWIR: the best accuracies were 0.96 for subsurface SWIR (GB; AUC = 0.99; MCC = 0.95) and 0.89 for surface SWIR (KNN; AUC = 0.98; MCC = 0.87). The sand and clay attributes had R² values ranging from 0.89 to 0.98. The prediction of soil organic carbon had R² values ranging from 0.66 to 0.85. The silt attribute performed the worst, with R² values ranging from 0.52 to 0.65. Overall, subsurface SWIR data provided the most reliable basis for taxonomic discrimination, while horizon classification was feasible but reflected gradual spectral transitions throughout the profile.

2025
Disertaciones
1
  • GABRIELA BISSOLI SILVA
  • Efetios de produtos biológicos no controle de Xanthomonas vasicola pv. vasculorum

  • Líder : DAURI JOSE TESSMANN
  • MIEMBROS DE LA BANCA :
  • DAURI JOSE TESSMANN
  • RUBIA DE OLIVEIRA MOLINA
  • WILLIAM MARIO DE CARVALHO NUNES
  • Data: 27-feb-2025
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The cultivation of corn in Paraná is fundamental for both local and national agriculture, being one
    of the main sources of income for producers. However, productivity can be compromised by bacterial
    streak, caused by Xanthomonas vasicola pv. vasculorum (Xvv), which generates economic losses and
    compromises food security. In light of this problem, this study aimed to evaluate the effect of
    biological products in disease management. Experiments were conducted at the UEM Bacteriology
    Laboratory and in a greenhouse. In vitro tests, isolate RL1 was grown on nutrient agar medium, and
    bacterial suspensions adjusted to 10⁸ CFU/mL were exposed to Bioimune®, Caravan®, Zigurat®,
    Kasumin®, and Paládio® products at four doses (0.5x, 1x, 2x, and 3x). Treatments were applied to
    paper discs on the bacteria, and the formation of inhibition halos was evaluated after 24 hours at
    28°C. None of the tested products formed inhibitory halos against Xvv, except for the pentabiotic
    (positive control), which showed consistent inhibition. In vivo tests, MG593PWU hybrid plants were
    grown in pots (5L) and at approximately 15 days old, were treated with the same products in six
    groups (negative control and five products), following the recommended doses for Xvv. After 72
    hours, the plants were inoculated with Xvv (10⁸ CFU/mL) in the whorl and maintained under high
    humidity for 15 hours. Disease severity was assessed at 9, 15, and 21 days, with data converted
    into Area Under the Disease Progress Curve (AUDPC). Statistical analysis revealed no significant
    difference (p ≤ 0.05) between biological and chemical treatments in reducing disease severity.
    Among the tested products, Kasumin®, Zigurat®, and Paládio® performed better in maintaining low
    severity over time, while Bioimune® and Caravan® had similar progressions to the control,
    indicating low efficacy in managing bacterial streak. Graphical analysis and logistic and Gompertz
    mathematical models reinforced the absence of statistical difference between treatments, suggesting
    that none of the evaluated products were significantly more effective than the control in reducing
    the disease.

2
  • ANDRESSA DA SILVA TINTI
  • Uso de medicamentos homeopáticos para o controle de mofo branco na cultura da alface

  • Líder : KATIA REGINA FREITAS SCHWAN ESTRADA
  • MIEMBROS DE LA BANCA :
  • BRUNO REIS
  • JOSE RENATO STANGARLIN
  • KATIA REGINA FREITAS SCHWAN ESTRADA
  • Data: 28-feb-2025
    Ata de defesa assinada:


  • Resumen Espectáculo
  • Lettuce (Lactuca sativa L.) is one of the most widely consumed vegetables in the world, but its productivity can be reduced by more than 50% due to white mold, caused by Sclerotinia sclerotiorum. Chemical control is the most common strategy for managing the disease, but due to the negative impacts, such as environmental contamination and health risks, there is a growing search for more sustainable cultivation systems. In this context, alternatives such as plant homeopathy have gained prominence. This study evaluated the potential of the homeopathic medicines Sulphur and Lycopodium clavatum, as well as biotherapeutics derived from the sclerotia and hyphae of the pathogen, in the 6CH, 12CH, 24CH, 30CH and 36CH dynamizations. The experiments included in vitro tests to verify the antimicrobial activity against S. sclerotiorum, and in vivo tests to analyze the interaction between lettuce and the pathogen, in addition to the effects on seed germination. Distilled water and a hydroalcoholic solution (70% alcohol) were used as controls. The results indicated that homeopathic medicines, incorporated into the PDA culture medium, influenced mycelial growth and sclerotia formation. Lycopodium significantly reduced fungal growth in all dynamizations, with emphasis on 30CH, which inhibited 91.83% of growth. Sulphur stimulated mycelial growth in some dynamizations. Biotherapeutics derived from hyphae and sclerotia also reduced fungal growth in all dynamizations tested. In sclerotia formation, Lycopodium was the most efficient in reducing the number, while the weight of sclerotia was lower with the Biotherapeutic Hifa 36CH. In treatments applied to the surface of the medium, Sulphur showed better performance, reducing mycelial growth and the number of sclerotia in all treatments, although without significant difference in weight. Carpogenic germination tests indicated that Lycopodium was the most effective. However, in the evaluation of disease progression and seed germination, there was no significant difference between treatments. The weight and dry mass of lettuce plants were positively influenced by the biotherapeutics Escleródio 36CH and Hifa 36CH. For healthy plants, the Biotherapeutic Escleródio 12CH was the most efficient treatment, evidencing its potential in promoting plant growth.

3
  • LUCAS RAFAEL SAPATINI DOS SANTOS
  • Utilização de sensores hiperespectrais para avaliação dos sintomas de deficiência de fósforo na cultura do tabaco (Nicoatiana tabacum)

  • Líder : MARCOS RAFAEL NANNI
  • MIEMBROS DE LA BANCA :
  • MARCOS RAFAEL NANNI
  • WERNER CAMARGOS ANTUNES
  • PETERSON RICARDO FIORIO
  • Data: 28-feb-2025
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The use of hyperspectral sensors has emerged as a promising tool for identifying and quantifying nutritional changes in plants, such as phosphorus deficiencies. However, despite technological advancements, the applicability of vegetation indices (IV’s) like NDVI still exhibits limitations under specific conditions, such as the detection of intermediate deficiencies. Hyperspectral sensors, when properly calibrated and processed, allow for greater precision in detecting subtle changes in the spectral response of plants. The aim of this study was to investigate the spectral behavior of tobacco plants subjected to different phosphorus levels in a nutrient solution, using a spectroradiometer and a HeadWall Photonics hyperspectral imaging sensor. The study sought to evaluate the ability of vegetation indices, such as NDVI, and to correlate spectral data with chemical and physical analyses of the plants to predict foliar phosphorus content. Predictive models, such as Partial Least Squares Regression (PLSR) and Random Forest (RF), were applied to estimate nutrient levels, considering the impact of phosphorus deficiencies on the visible and near-infrared spectrum.The experiment was conducted in a randomized block design (RBD) with 11 treatments and 10 replicates, varying phosphorus concentrations from 0% to 100%. Data collection was carried out after 22 days of hydroponic cultivation, including spectral readings, chemical analyses, and measurements of vegetation indices such as NDVI, CARI, CVI, and REIP. The results showed significant changes in spectral behavior between treatments, particularly in leaves with phosphorus deficiencies, highlighting differences in the red and near-infrared regions. While NDVI proved effective in identifying severe deficiencies, it was unable to adequately differentiate intermediate conditions. On the other hand, PLSR-based models demonstrated greater accuracy in predicting phosphorus content, outperforming Random Forest and indicating potential for rapid, non-destructive diagnostics

4
  • ANA CLARA PEREIRA DE SOUZA VERGINIO
  • Eficiência agronômica de fertilizantes organominerais em milho em Latossolo Vermelho distroférrico.

  • Líder : MARCELO AUGUSTO BATISTA
  • MIEMBROS DE LA BANCA :
  • MARCELO AUGUSTO BATISTA
  • ADÔNIS MOREIRA
  • MARCELO ALESSANDRO ARAUJO
  • Data: 28-mar-2025
    Ata de defesa assinada:


  • Resumen Espectáculo
  • The cost of phosphate fertilizers, the increasing demand, and the progressive depletion of
    reserves have led farmers, agronomists, and researchers to seek more efficient use of this
    non-renewable resource. One alternative is the use of organomineral fertilizers, which result from
    the combination of organic and mineral fertilizers. This study aimed to evaluate the performance of
    different phosphate fertilizers (organomineral and mineral) and application rates (standard
    fertilization and reduced dose) in a first-season maize crop. The organomineral fertilizers used
    were manufactured from different organic materials and processes. The organic materials included
    peat, poultry litter, and composting, while the manufacturing processes comprised granulation,
    granular mixing, and pelletization. The fertilizers tested were: mineral granular mix (MIN),
    granulated organomineral with peat (OG), granular mix organomineral with poultry litter (OMG), and
    pelletized organomineral with composting (OPE). The 100% and 50% of the standard recommendation
    corresponded to 30 and 15 kg ha⁻¹ of P₂O₅, respectively. The nitrogen (N) from urea + NBPT and
    potassium (K₂O) from KCl were adjusted to match all treatments. The experiment was conducted on an
    Oxisol in Floresta-PR during the 2023/24 growing season. The analyzed variables included yield,
    thousand-grain weight, P content in leaves, P content in grains, total P exported, partial
    productivity factor index (PPF), and partial P balance (PPB), soil P, organic carbon (OC), as well
    as the correlation between these variables. Organomineral fertilizers showed performance similar to
    mineral fertilizers in supplying P to plants. Higher yields were achieved with higher fertilizer
    doses, regardless of the source used. The OMG fertilizer outperformed OPE in terms of soil P, P
    export, PPF, and PPB, indicating that the organic raw material and the production method of
    organomineral fertilizers influence their field performance. However, organomineral fertilizers did
    not directly or indirectly increase the organic carbon content in the soil.

Tesis
1
  • AILTON DONIZETE SILVÉRIO
  • -

  • Líder : PAULO SERGIO LOURENCO DE FREITAS
  • MIEMBROS DE LA BANCA :
  • CARLOS SÉRGIO TIRITAN
  • GISELE DAIANE PINHA
  • PAULO SERGIO LOURENCO DE FREITAS
  • ROBERTO REZENDE
  • TELMO ANTONIO TONIN
  • Data: 09-dic-2025


  • Resumen Espectáculo
  • -

2024
Disertaciones
1
  • FERNANDA RITA CORREIA
  • Uso da susceptibilidade magnética e da compoisção elemental na discriminação forense de amostras de terra preta de índio e argissolo

  • Líder : ANTONIO CARLOS SARAIVA DA COSTA
  • MIEMBROS DE LA BANCA :
  • ANTONIO CARLOS SARAIVA DA COSTA
  • IVAN GRANEMANN DE SOUZA JUNIOR
  • RODRIGO STUDART CORREA
  • Data: 29-ago-2024
    Ata de defesa assinada:


  • Resumen Espectáculo
  • This study aims to analyze, through the magnetic susceptibility method, the magnetic activity of
    soil samples collected every 10 cm up to 100 cm in depth and in three sampling transects (A, B, C)
    established in the southwest-northeast direction, kept 100 m apart from each other, with ten soil
    sampling points allocated every 10 m along each transect and with points 1 to 5 referring to the
    Terra Preta de Índio (TPI) soil and points 6 to 10 to the Argisol collected in the Caldeirão
    Experimental Field, which is located in Contagem de Iranduba (3º15'6.8'' South and 60º13'46.9''
    West - WGS 84 datum), 20 m above the bank of the Solimões River, Amazon River sedimentary basin,
    Tertiary Amazonian Plateau, Western Amazon, Brazil, from a forensic point of view. evaluating the
    magnetic activity at the surface and in the deeper layers of the soil and verifying whether the SM
    analysis is capable of differentiating both vertically and horizontally the samples. The elemental
    analysis was conducted according to the article “Elemental signatures of an Amazonian Dark Earth as
    result of its formation process”. Elemental analyzes of Ag - silver, Al - aluminum, As - arsenic,
    Ba - barium, Be - beryllium, Ca- calcium, Cd - cadmium, Co - cobalt, Cr - chromium, Cs - cesium, Cu - copper, Fe - iron, K -potassium, Li - lithium, Mg - magnesium, Mn - manganese, Mo - molybdenum, Ni - nickel, P
    - phosphorus, Pb - lead, Rb - rubidium, Se - selenium, Sr - strontium, Tl - thallium, U - uranium,
    V - vanadium and Zn - zinc were performed and were conducted on an ICP-MS Agilent 7500cx. The
    statistical software used was SPSS Statistics and was chosen based on the specific requirements of
    the study and the type of statistical tests needed to answer the research questions. The results of
    the statistical analysis were interpreted in the context of the research objectives and used to
    draw conclusions and make recommendations for future studies or applications of the findings.

2
  • FRANCIELI CRISTINA GESSI
  • -

  • Líder : KATIA REGINA FREITAS SCHWAN ESTRADA
  • MIEMBROS DE LA BANCA :
  • FRANCIELLI GASPAROTTO
  • HIGO FORLAN AMARAL
  • KATIA REGINA FREITAS SCHWAN ESTRADA
  • Data: 30-ago-2024


  • Resumen Espectáculo
  • -

2023
Tesis
1
  • MARCO DOS SANTOS MARTINEZ
  • .

  • Líder : ANTONIO CARLOS SARAIVA DA COSTA
  • MIEMBROS DE LA BANCA :
  • ANIBAL DE MORAES
  • ANTONIO CARLOS SARAIVA DA COSTA
  • FABIO CORTEZ LEITE DE OLIVEIRA
  • HENRIQUE LEAL PEREZ
  • MARCOS WEBER DO CANTO
  • Data: 28-ago-2023


  • Resumen Espectáculo
  • .

SIGAA | DAA - Pós-Graduação - (44) 3011-4475 | Copyright © 2006-2026 - UEM - A9-SigaaPA3.srv3inst1