https://repositorio.ufra.edu.br/jspui/handle/123456789/18 Programa de Pós-Graduação em Agronomia Mon, 28 Sep 2026 21:22:47 GMT 2026-09-28T21:22:47Z http://repositorio.ufra.edu.br:443/jspui/retrieve/ae3b45ba-2dc7-44cc-890e-9f8117a1bd99/logo AGRONOMIA.png https://repositorio.ufra.edu.br/jspui/handle/123456789/18 https://repositorio.ufra.edu.br/jspui/handle/riufra/2905 Título: AI-driven digital soil Mapping of the Eastern Amazon: integrating pedological knowledge and pedometrics Autor(es): SOBRINHO, Rômulo José Alencar Abstract: The lack of more detailed information on soil distribution in the Amazon is an impediment to effective soil governance in this region. Therefore, the overall objective of this study was to evaluate the performance and uncertainty of assembly algorithm approaches, legacy mapping, environmental covariates, and reference area. To this end, this thesis was divided into two chapters, titled: I) Integrating Tacit Knowledge and AI for Digital Soil Mapping in Eastern Amazonia: Ensemble Learning, Model Performance, and Uncertainty Incorporation; and II) Transferring soil–landscape relationships in data-limited regions of the Amazon: a knowledgeguided digital soil mapping approach based on reference areas. Output data from both chapters followed the common steps below. Fifteen environmental covariates based on the SCORPAN model were used. To generate the predictive models, we used the algorithms and four input datasets: Random Forest, Ranger, Xgboost, C 5.0, and Ensemble Learning. We also evaluated two levels of detail for the Mapping Units of the legacy map: taxonomic order level and large group level. Terrain covariates were derived from two distinct elevation models: digital surface model and digital terrain model. A total of 20 predictive models were analyzed. Key findings from Chapter 2: the algorithm assembly approach using a digital surface model and unit mapping at the Large Group taxonomic level achieved statistically superior performance in terms of accuracy and Kappa compared to individual models at the Large Group taxonomic level. Environmental covariates such as: 22,000-year-old Mean Annual Temperature, Channel Network Base Level, Altitude, and Land Surface Temperature showed greater importance in predictive modeling; And as highlights of Chapter 3: in the external validation in the extrapolation area, the ensemble learning predictive models using data derived from the digital surface model, in mapping units at the Oreder taxonomic level, ensemble learning using data derived from the digital terrain model, in mapping units at the Order taxonomic level and ensemble learning using data derived from the digital surface model, in mapping units at the Great group taxonomic level, and the prediction of the mapping units GLEISSOLOS and GLEISSOLOS SÁLICOS Sódicos stood out as the most accurate in the extrapolation area. The maps generated from the legacy map and for the extrapolation area, using reference area, successfully mapped mapping units related to the hydromorphic environment. The pioneering nature of this study in the Amazon revealed the potential of the ensemble learning approach for mapping soil classes under hydromorphic environments, environments that occupy a large area in the Amazon and are of significant importance.; The lack of more detailed information on soil distribution in the Amazon is an impediment to effective soil governance in this region. Therefore, the overall objective of this study was to evaluate the performance and uncertainty of assembly algorithm approaches, legacy mapping, environmental covariates, and reference area. To this end, this thesis was divided into two chapters, titled: I) Integrating Tacit Knowledge and AI for Digital Soil Mapping in Eastern Amazonia: Ensemble Learning, Model Performance, and Uncertainty Incorporation; and II) Transferring soil–landscape relationships in data-limited regions of the Amazon: a knowledgeguided digital soil mapping approach based on reference areas. Output data from both chapters followed the common steps below. Fifteen environmental covariates based on the SCORPAN model were used. To generate the predictive models, we used the algorithms and four input datasets: Random Forest, Ranger, Xgboost, C 5.0, and Ensemble Learning. We also evaluated two levels of detail for the Mapping Units of the legacy map: taxonomic order level and large group level. Terrain covariates were derived from two distinct elevation models: digital surface model and digital terrain model. A total of 20 predictive models were analyzed. Key findings from Chapter 2: the algorithm assembly approach using a digital surface model and unit mapping at the Large Group taxonomic level achieved statistically superior performance in terms of accuracy and Kappa compared to individual models at the Large Group taxonomic level. Environmental covariates such as: 22,000-year-old Mean Annual Temperature, Channel Network Base Level, Altitude, and Land Surface Temperature showed greater importance in predictive modeling; And as highlights of Chapter 3: in the external validation in the extrapolation area, the ensemble learning predictive models using data derived from the digital surface model, in mapping units at the Oreder taxonomic level, ensemble learning using data derived from the digital terrain model, in mapping units at the Order taxonomic level and ensemble learning using data derived from the digital surface model, in mapping units at the Great group taxonomic level, and the prediction of the mapping units GLEISSOLOS and GLEISSOLOS SÁLICOS Sódicos stood out as the most accurate in the extrapolation area. The maps generated from the legacy map and for the extrapolation area, using reference area, successfully mapped mapping units related to the hydromorphic environment. The pioneering nature of this study in the Amazon revealed the potential of the ensemble learning approach for mapping soil classes under hydromorphic environments, environments that occupy a large area in the Amazon and are of significant importance.; The lack of more detailed information on soil distribution in the Amazon is an impediment to effective soil governance in this region. Therefore, the overall objective of this study was to evaluate the performance and uncertainty of assembly algorithm approaches, legacy mapping, environmental covariates, and reference area. To this end, this thesis was divided into two chapters, titled: I) Integrating Tacit Knowledge and AI for Digital Soil Mapping in Eastern Amazonia: Ensemble Learning, Model Performance, and Uncertainty Incorporation; and II) Transferring soil–landscape relationships in data-limited regions of the Amazon: a knowledgeguided digital soil mapping approach based on reference areas. Output data from both chapters followed the common steps below. Fifteen environmental covariates based on the SCORPAN model were used. To generate the predictive models, we used the algorithms and four input datasets: Random Forest, Ranger, Xgboost, C 5.0, and Ensemble Learning. We also evaluated two levels of detail for the Mapping Units of the legacy map: taxonomic order level and large group level. Terrain covariates were derived from two distinct elevation models: digital surface model and digital terrain model. A total of 20 predictive models were analyzed. Key findings from Chapter 2: the algorithm assembly approach using a digital surface model and unit mapping at the Large Group taxonomic level achieved statistically superior performance in terms of accuracy and Kappa compared to individual models at the Large Group taxonomic level. Environmental covariates such as: 22,000-year-old Mean Annual Temperature, Channel Network Base Level, Altitude, and Land Surface Temperature showed greater importance in predictive modeling; And as highlights of Chapter 3: in the external validation in the extrapolation area, the ensemble learning predictive models using data derived from the digital surface model, in mapping units at the Oreder taxonomic level, ensemble learning using data derived from the digital terrain model, in mapping units at the Order taxonomic level and ensemble learning using data derived from the digital surface model, in mapping units at the Great group taxonomic level, and the prediction of the mapping units GLEISSOLOS and GLEISSOLOS SÁLICOS Sódicos stood out as the most accurate in the extrapolation area. The maps generated from the legacy map and for the extrapolation area, using reference area, successfully mapped mapping units related to the hydromorphic environment. The pioneering nature of this study in the Amazon revealed the potential of the ensemble learning approach for mapping soil classes under hydromorphic environments, environments that occupy a large area in the Amazon and are of significant importance. Tue, 30 Jun 2026 00:00:00 GMT https://repositorio.ufra.edu.br/jspui/handle/riufra/2905 2026-06-30T00:00:00Z https://repositorio.ufra.edu.br/jspui/handle/riufra/2901 Título: Sistema integrado de diagnose e recomendação (DRIS) na avaliação do estado nutricional da cultura do coqueiro hibrido Autor(es): SALDANHA, Eduardo Cézar Medeiros Abstract: Among the methods used for the interpretation of the results of foliar analysis, there is the integrated system of diagnosis and recommendation (DRIS). This method is based on the caleulation of an index for each nutrient. For the calceulation of DRIS norms, there is a need to organize a database of foliar orchards with known productivity, and establish standards for the relationships between nutrients. In the Amazonian conditions, specifically in the state of Pará, no DRIS norms developed for hybrid coconut culture. Hus the aim of this work was to develop DRIS norms for different selection criteria of the relationships between nutrients for growing the hybrid coconut in Pará. The study was conducted in the municipality of Farm SOCOCO Moju PA. Was used to form the database results of leaf analysis and productivity of 134 observations for the period 2001-2011. Obtained the mean, standard deviation, coefficient of variation and variance of the relationships of the nutrients, the leaf samples of vintages from 2001 to 2011, the coefficient of correlation between the ratio of each pair of nutrients and fruit yield. The highest values of standard deviation, variance and coefficient of variation were presented for foliar concentrations of the micronutrients iron, manganese and boron. The nutrients showed higher percentages of samples with levels below the appropriate levels used were Mg and Ca, while nutrients showed that samples with foliar concentrations above or equal to the appropriate levels were Fe, Mn and Ca. DRIS norms for growing hybrid greens were established, based on the relationships between nutrients in the population of high productivity. Of the 110 relationships between nutrients studied, 55 relationships were selected to compose the DRIS norms for the cultivation of hybrid coconut, using two selection criteria of the relationships between nutrients.; The assessment of the nutritional status of plants by leaf analysis as the Integrated System Diagnosis and Recommendation (DRIS) has been highlighting the traditional methods of interpreting the results of the analysis of plant tissue. From the DRIS indices expressing the balance of nutrients in a plant, comparing relationships in the sample being diagnosed with rules or default values are calculated. For the coconut, nutritional counseling through foliar analysis is touted as an efficient method for fertilizer recommendation, and the results have been traditionally interpreted using the criteria and the critical level of sufficiency ranges. The objective of this study was to evaluate the nutritional status and establish nutritional standards for the cultivation of hybrid coconut in the municipality of Moju, Pará, using the DRIS. Was used to form the database results of leaf analysis and productivity of 134 observations for the period 2001-2011. To calculate DRIS indices of 134 leaf samples were used DRIS norms established in the first chapter of this work. It was found for the most common deficiencies element K, and Mg possibly in excess element. The order of nutrient limitation, was K> P>Ca> Fe> N> O> B> Zn> Cu> Mn> Mg. The Ca, Fe and K nutrients are more likely to respond positively to fertilization, since the Mg, Cu and Mn nutrients, were diagnosed as having the greatest like lihood of negative response to fertilization. It was also found that N and P are the ones that are in the best position to nutritional balance. Regression equations for the relationship between the nutrient content in leaves of hybrid coconut and its DRIS indices, which allowed establishing nutrient reference values, based on DRIS were adjusted.; The DRIS method has been used infrequently as a tool for evaluating theresults of fertilization experiments in different crops. Watts work has shown that the method has been able to identify nutritional limitations with different culture, thus confirming its robustness and efficiency in nutritional diagnostics. The aim of this study carry the nutritional diagnosis of the culture of hybrid coconut, from the use of the critical level method and the integrated system of diagnosis and recommendation (DRIS) on the results of experiment evaluating different sources of phosphorus. Nutritional diagnostics for macro and micronutrients were drawn from critical levels (CL) and integrated system of diagnosis and recommendation (DRIS), with results obtained from the experiment of phosphorus fertilization on crops in 2004, 2005, 2006 and 2007 data and productivity levels leaf were used from this experiment in the area of commercial coconut production company Socôco - Agribusiness Amazon. There was no significant response to the productivity variables assessed in any of the four years evaluated. Despite the absence of significant effects, it was found that the treatments resulted in different values of productivity. It was found that the nutritional diagnostic prepared from the DRIS method of predicting were consistent fills or nutritional deficiencies for P element, depending on the treatments to different sources and combinations, since with reference to the traditional method of interpreting through the level critical leaf, there was agreement in the nutritional information for P. Tue, 26 May 2026 00:00:00 GMT https://repositorio.ufra.edu.br/jspui/handle/riufra/2901 2026-05-26T00:00:00Z https://repositorio.ufra.edu.br/jspui/handle/riufra/2886 Título: Influência da cobertura pedológica na utilização do solo na localidade de Benfica, município de Itupiranga, PA. Autor(es): SIMÔES, Lourdes Henchen Ritter Abstract: L'étude du sol dans le contexte de Tagriculture familiale en front pionnier est importante pour mettre en évidence les correlations entre les systèmes pédologiques, leurs potentiels et leurs limitations et les modes d'utilisation agricole locale. La principale fmalité de ce travail est d'identifíer la diversité des types de sois, leurs aptitudes culturales et comment ils sont utiiisés par les agriculteurs, à partir d'observations morphologiques sur trois toposséquences de sois formées sur deux roches (monzogranite et granodiorite), et d^ntretiens avec les agriculteurs Socaux. Les résultats montrent quMl existe une diversité de sois avec des potentiels agronomiques variés, offrant des alternatives durables pour Tagriculture et le maintien des exploitations. Cette diversité de systèmes pédologiques dans Ia plupart des cas n/est nas prise en compte par les agriculteurs. Cependant cette attitude est liée à une stratégie de production bovine dans laquelle les intérêts paraissent être la valorisation du lot en pâturage pour une vente future. Toutefois, quelques agriculteurs adoptent des stratégies dê diversificâtion du système de production et, alors, les différents types de sois sont considérés. Dans ce cas rexpioitation devient plus viable, garantissant pour une plus grande période le maintien de la capacité productive du milieu. Mon, 11 May 2026 00:00:00 GMT https://repositorio.ufra.edu.br/jspui/handle/riufra/2886 2026-05-11T00:00:00Z https://repositorio.ufra.edu.br/jspui/handle/riufra/2861 Título: Distribuição espacial e amostragem sequencial de Spodoptera frugiperda (J.E. Smith, 1797) na cultura do milho. Autor(es): FARIAS, Paulo Roberto Silva Abstract: With the aim of studying the distribution of Spodoptera frugiperda (J.E. Smith, 1797) in corn crops, a stratified sampling system was implemented, consisting of three fields, each comprising 100 plots, conducted in a corn field at the Faculty of Agricultural and Veterinary Sciences in Jaboticabal during the 1994/1995 growing season. The number of caterpillars per plant was counted on 16 sampling dates. The data obtained were fitted to the negative binomial, Poisson, and positive binomial distributions. The following aggregation indices were also studied: dispersion index (s/m), Morisita index, |d statistic, K parameter of the negative binomial distribution, and the b coefficient of Taylor’s law. The number of caterpillars per plant was counted on 16 sampling dates. The data obtained were fitted to negative binomial, Poisson, and positive binomial distributions. The following aggregation indices were also studied: dispersion index (s/m), Morisita index, |d statistic, K parameter of the negative binomial distribution, and the b coefficient of Taylor’s law. The studies revealed that the numbers of small caterpillars per plant fit the negative binomial distribution, while the numbers of large caterpillars exhibited a distribution tending toward randomness. The aggregation indices showed a marked tendency toward aggregated spatial distribution for small, large, and total S. frugiperda. Sequential sampling plans were constructed based on the number of caterpillars per plant and on the percentage of infested plants. Tue, 31 Mar 2026 00:00:00 GMT https://repositorio.ufra.edu.br/jspui/handle/riufra/2861 2026-03-31T00:00:00Z