Abstract Artificial neural networks (ANNs) may experience problems due to insufficient or uninformative predictors, and these problems are common for complex predictions such as those for rainfall. However, some studies point to the use of climate variables and anomalies as predictors to make the forecast more accurate. This research aimed to predict the monthly rainfall, one month in advance, in four municipalities of the metropolitan region of Belo Horizonte using an ANN trained with different climate variables; additionally, it aimed to indicate the suitability of such variables as inputs to these models. The models were developed using the MATLAB® software Version R2011a using the NNTOOL toolbox. The ANNs were trained by the multilayer perceptron architecture and the feedforward and backpropagation algorithm using two combinations of input data, with two and six variables, and one combination of input data with the three most correlated variables to observed rainfall from 1970 to 1999 to predict the rainfall from 2000 to 2009. The climate variable most correlated with the rainfall of the following month was the average compensated temperature. Even when using the variables most correlated with precipitation as predictors (0.66 ≤ nt index ≤ 1.26), there was no notable improvement in the predictive capacity of the models when compared to those that did not use climate variables as predictors (0.55 ≤ nt index ≤ 0.80). (ANNs However accurate advance additionally MATLAB Ra R a toolbox 197 199 200 2009 temperature 0.66 066 0 66 (0.6 1.26, 126 1.26 , 1 26 1.26) 0.55 055 55 (0.5 0.80. 080 0.80 . 80 0.80) 19 20 0.6 06 6 (0. 12 1.2 2 0.5 05 5 08 0.8 8 0. (0 1. (

Resumo As redes neurais artificiais (RNAs) podem apresentar problemas devido a preditores insuficientes ou não informativos, o que é comum para previsões complexas, como as de precipitação pluvial. No entanto, alguns estudos apontam para o uso de variáveis e anomalias climáticas como preditores para tornar a previsão mais precisa. Esta pesquisa teve como objetivo prever a precipitação mensal, com um mês de antecedência, em quatro municípios da região metropolitana de Belo Horizonte utilizando uma RNA treinada com diferentes variáveis climáticas; além disso, buscou indicar a adequação de tais variáveis como entrada para esses modelos. Os modelos foram desenvolvidos por meio do software MATLAB® versão R2011a utilizando a toolbox NNTOOL. As RNAs foram treinadas pela arquitetura multilayer perceptron e pelo algoritmo feedforward e backpropagation usando duas combinações de dados de entrada, com duas e seis variáveis, e uma combinação de dados de entrada com as três variáveis mais correlacionadas com precipitação observada de 1970 a 1999 para prever a precipitação de 2000 a 2009. A variável climática mais correlacionada com a precipitação do mês seguinte foi a temperatura média compensada. Mesmo utilizando as variáveis mais correlacionadas com a precipitação como preditores (0,66 ≤ índice nt ≤ 1,26), não houve melhora significativa na capacidade preditiva dos modelos quando comparado aos que não utilizaram variáveis climáticas como preditores (0,55 ≤ índice nt ≤ 0,80). (RNAs informativos complexas pluvial entanto precisa mensal antecedência disso MATLAB Ra R NNTOOL 197 199 200 2009 compensada 0,66 066 0 66 (0,6 1,26, 126 1,26 , 1 26 1,26) 0,55 055 55 (0,5 0,80. 080 0,80 . 80 0,80) 19 20 0,6 06 6 (0, 12 1,2 2 0,5 05 5 08 0,8 8 0, (0 1, (

RESUMO A temperatura do ar afeta significativamente os processos que envolvem atividades agrícolas e humanas. O conhecimento da temperatura de um determinado local é fundamental para o planejamento agrícola. Também ajuda a tomar decisões sobre as atividades humanas. No entanto, nem sempre é possível determinar essa variável. É necessário fazer uma estimativa precisa, utilizando métodos que sejam capazes de detectar as variações existentes. O objetivo deste estudo foi desenvolver modelos de regressão linear múltipla (RLM), rede neural artificial (RNA) e floresta aleatória (FA) para estimar a temperatura média (Tmean), máximo (Tmax), e mínimo (Tmin) mensal do ar em função de coordenadas geográficas e altitude para diferentes áreas do Estado de Minas Gerais, Brasil, com classificação climática Cwa ou Cwb. Os dados médios mensais (Tmean, Tmax e Tmin), ao longo de um período de 30 anos, foram coletados em 20 estações climatológicas. O RLM foi capaz de estimar o Tmax com precisão. Porém, a capacidade preditiva de estimar Tmean e Tmin foi baixa. Os algoritmos FA e RNA foram usados para estimar Tmean, Tmax e Tmin com alta precisão. Os melhores resultados foram obtidos com o modelo RF.

ABSTRACT Air temperature significantly affects the processes involving agricultural and human activities. The knowledge of the temperature of a given location is essential for agricultural planning. It also helps to make decisions regarding human activities. However, it is not always possible to determine this variable. It is necessary to make a precise estimate, using methods that are capable of detecting the existing variations. The aim of this study was to develop models of multiple linear regression (MLR), artificial neural network (ANN), and random forest (RF) to estimate the mean (Tmean), maximum (Tmax), and minimum (Tmin) monthly air temperatures as a function of geographic coordinates and altitude for different localities in Minas Gerais state, Brazil, with climatic classification Cwa or Cwb. The average monthly data (Tmean, Tmax, and Tmin), over a period of 30 years, were collected from 20 climatological stations. The MLR was able to estimate the Tmax with accuracy. However, the predictive capacity of estimating Tmean and Tmin was low. The algorithms RF and ANN were used to estimate Tmean, Tmax, and Tmin with high accuracy. The best results were obtained using the RF model.

ABSTRACT Agriculture has caused numerous concerns regarding the preservation of natural resources. In this context, agroforestry systems are emerging as a more sustainable alternative. The present study aimed to evaluate growth characteristics, radiation use efficiency, biomass partition, and yield of soybean grown in two agroforestry systems and full sun. A field experiment was conducted in Southern Brazil during the 2014/2015 crop year, in which transmissivity of solar radiation, leaf area index, radiation use efficiency, and yield of soybean were evaluated. The solar radiation dynamics varied among the study factors, mainly due to the agroforestry arrangements. Shading influenced the leaf area index, radiation use efficiency, biomass partition, and soybean yield. Based on the soybean yield values generated in this study, the use of more spaced agroforestry arrangements and Peltophorum dubium forest species is recommended. However, because the yield values were below the expected levels, it is not yet possible to confirm the full potential of soybean crop in agroforestry systems. Thus, new studies should be conducted in order to generate alternatives that make soybean cultivation feasible in agroforestry systems, such as assessing the use of more spaced agroforestry arrangements, the insertion of the soybean crop in the initial years of cultivation of agroforestry, as well as reduce intraspecific competition by decreasing the plant population of the crop.

ABSTRACT Aloysia triphylla is an aromatic plant used in several industrial sectors, owing to the chemical properties of its essential oil. Biosynthesis of organic compounds is influenced by the growth conditions. As such, temporal and spatial variation affect the chemical composition of essential oils. We hypothesized that: i) the chemical composition of A. triphylla essential oil is influenced by differences in irrigation and season; and ii) the major components of A. triphylla essential oil do not vary between treatment conditions. To test these hypotheses, we determined the chemical composition of A. triphylla essential oil as a function of four seasons crossed with four irrigation levels. A completely randomized experimental design with a randomized block in a 4 × 4 factorial scheme, representing the four seasons of the year (summer, autumn, winter, and spring) and four irrigation levels (50%, 75%, 100%, and 125% of the reference evapotranspiration), was used with four replicates. Our results show that the chemical composition of A. triphylla essential oil varied with the two study factors, of which season was the major factor. The highest concentrations of constituents classified as monoterpene and sesquiterpene were observed in the summer season, and the highest constituents of the other group of compounds were observed in winter. In addition, the major components of the essential oil were α-citral, limonene, and β-citral, and their levels were the highest during winter.

Abstract: The objective of the study was to evaluate the growth and production efficiency of forage-grain dual-purpose wheat in two arrangements and four agroforestry systems (Intercrop-I and Intercrop-II), as well as to evaluate crop management for the wheat crops in two planting seasons (Crop Season I-2014 and Crop Season II-2015. The experiment was conducted in a randomized complete block design, factorial scheme 7x2x2, with seven cultivation systems Eucalyptus urophylla x Eucalyptus grandis Intercrop-II and Intercrop-I; Peltophorum dubium Intercrop-II and Intercrop-I; P. rigida Intercrop-I; S. parahyba Intercrop-I; a wheat monoculture with no tree species present; and two harvest management techniques (with and without harvesting of the forage species). Agroforestry systems generated shading for wheat plants, with a higher phyllochron and lower values of leaf area index of those individuals kept under trees with higher crown shading (non-deciduous trees) due to the lower transmissivity of solar radiation. The systems composed with Schizolobium parahyba in Intercrop-I and Parapiptadenia rigida in Intercrop-I provide a minor phyllochron to the wheat, resulting in a higher leaf area index and dry matter yield. Thus, the cultivation of tree species and dual-skilled agricultural crops, such as wheat, provides promising alternatives for the future use of land in tropical countries.

Abstract: Despite its economic importance, the microclimate in soybean canopies has not yet been studied in detail. Such a study can yield valuable information regarding the interaction of a crop with its environment. In this context, the aim of this study was to evaluate the solar radiation dynamic and yield responses for each canopy strata for two soybean cultivars with determined and undetermined growth habits. A field study was conducted during the 2013/2014 and 2014/2015 growing seasons in the city of Frederico Westphalen, Rio Grande do Sul, Brazil. The cultivar NA6411, with a determinate growth habit, presented a greater interception of radiation in the middle and lower canopies strata which results in higher soybean yield when compared to the cultivar TEC6029, and thus, can be recommended for cultivation in crop production systems. The contribution of the middle and upper canopy strata to the total yield formation was greater than that observed in the lower canopy strata due to the greater interception of solar radiation by these strata. To increase soybean yields, new studies regarding the microclimatic conditions of the soybean canopy should be developed to improve the maximum potential yield of the new soybean cultivars.

Resumo O objetivo deste trabalho foi determinar o crescimento, a eficiência do uso da radiação solar e a proteína bruta do sorgo forrageiro submetido a diferentes manejos de cortes. Foi utilizado um esquema unifatorial em delineamento de blocos casualizados (manejo de cortes), que variou do tratamento sem corte até quatro cortes. O primeiro corte ocorreu 44 dias após a semeadura, o segundo 23 dias após o rebrote, o terceiro 25 dias após o rebrote e o último após 40 dias do rebrote. A eficiência do uso da radiação solar do sorgo forrageiro reduziu com o aumento do número de cortes. A proteína bruta das plantas de sorgo aumentou com o manejo de cortes. Portanto, para a cultura do sorgo forrageiro, recomenda-se restringir o manejo de cortes para até três vezes durante o ciclo, desde que exista potencial produtivo de biomassa, e esses cortes devem ocorrer a cada 25 dias, dependendo do acréscimo de matéria seca. Não se recomenda a realização de mais do que três cortes, devido à baixa eficiência na utilização da radiação solar, à reduzida acumulação de matéria seca e à baixa qualidade desse material.

Abstract The purpose of this study is to determine the growth, the solar radiation use efficiency, and the gross protein of sorghum forage under different cutting management. A unifactorial design in random blocks (cutting management) was used, varying from no-cut treatment to four cuts. The first cut occurred 44 days after sowing, the second one after 23 days of regrowth, the third one after 25 days of regrowth, and the last one after 40 days of regrowth. The efficiency of the sorghum forage's solar radiation use reduces as the number of cuts increases. The sorghum forage's gross protein increases with cutting management. Therefore, for a sorghum forage crop, it is recommended to restrain the cutting management up to three times during the crop cycle, since it has the biomass potential production, and the cuts should be made at every 25 days, depending on the increase of dry mass. After that, the cutting management should not be done, mainly due to the low radiation use efficiency, reduced dry mass accumulation, and low material quality.

Abstract Sugarcane (Sacharum officinarum L.) monocropping has had a great socio-economic and environmental impact in Brazil, and agroforestry systems have been considered as an alternative for more sustainable production; however, there is a lack of field research under such conditions. The aim of this study was to evaluate the growth rates, radiation use efficiency and yield traits in sugarcane cultivated in the understory of Aleurites fordii, in two agroforestry arrangements and monocropping systems. A field experiment was conducted from July 2015 to June 2016 in the city of Frederico Westphalen, Rio Grande do Sul, Brazil. The radiation use efficiency, assimilate partitioning, leaf area index, absolute growth rate, net assimilation rate, number of tillers, plant height, % of intercepted solar radiation, extinction coefficient, and yield in each system was evaluated. In agroforestry systems, the dynamic interactions between multiple plant species change with the time and can result in unique microclimates. The use of agroforestry systems in 12 x 12m arrangements should be prioritized because it enables greater yields and radiation availability in the understory. This study sought to provide new sustainable alternatives for farmers in order to increase the diversification of the rural property and maintain the preservation of existing agroecosystems.

Abstract Soybean (Glycine max L.) monocropping have had a great socio-economic and environmental impact on the world and agroforestry systems has been considered an alternative for more sustainable production. The aim of this study was to evaluate the effect of different arrangements of agroforestry systems of the species E. urophylla x E. grandis and P. dubium and the pruning of the trees on the dynamics of solar radiation and soybean yield. A field experiment was conducted in Southern Brazil. In order to achieve the objective of the study, the solar radiation transmissivity and the productive performance of the soybean were evaluated. In this study, the soybean yield was determined by the arrangement of the agroforestry system and forest species utilized. The soybean crop cultivated in the understory of the P. dubium trees showed the highest yield response. The use of silvicultural practices for the management of forest species should be considered to generate the balance of the productive system. In this context, tree pruning is a viable side for agroforestry systems, as it promotes an increase in the incidence of solar radiation in the understory, resulting in an increase in soybean yield, especially when integrated with P. dubium.

ABSTRACT The objective of this experiment was to determine the conversion efficiency of intercepted photosynthetically active radiation into biomass of Acacia mearnsii De Wild. seedlings. A forest species, plastic tubes (90 cm3), and 11 evaluation periods (up to 180 days after emergence) were used in this study. The leaf area index (LAI), total dry biomass (BIO), global solar radiation (GSR), cumulative intercepted photosynthetically active radiation (PARic), and conversion efficiency of radiation (εb) were determined using a pyranometer (LI200X, LICOR). The value of εb in BIO seedlings of Acacia mearnsii was 7.76 g MJ-1. LAI was directly related to the efficiency of PARic, and this influenced the development, production potential and accumulation of BIO. The value of GSR flow was 11.81 MJ m-2 day-1, while the value inside the greenhouse was 6.26 MJ m-2 day-1.

RESUMO Esse trabalho teve como objetivo determinar o efeito da idade e do espaçamento de plantio na produção de biomassa das espécies florestais: Acacia mearnsii De Wild, Eucalyptus grandis W. Hill ex Maiden, Mimosa scabrella Benth e Ateleia glazioviana Baill. Foram considerados plantios com diferentes espaçamentos: 2,0×1,0 m, 2,0×1,5 m, 3,0×1,0 m e 3,0×1,5 m, nas idades de 1, 3 e 5 anos após o plantio. O trabalho foi instalado em delineamento experimental de blocos completos casualizados. A determinação da biomassa aérea consistiu na colheita e pesagem das diferentes partes das árvores. Os espaçamentos de plantio influenciam na produção de biomassa das partes das espécies florestais, nas diferentes idades após o plantio. Os espaçamentos mais adensados proporcionam maiores quantidades de biomassa que os menos adensados. As espécies florestais diferenciam-se em relação à produção de biomassa. O Eucalyptus grandis (325,1 t ha-1) apresenta a maior produção, seguido da Acacia mearnsii (239,3 t ha-1), Mimosa scabrella (53,4 t ha-1) e Ateleia glazioviana (32,1 t ha-1). A madeira apresenta a maior participação na biomassa total ao longo do tempo, sendo constatada a crescente proporção percentual ao longo das classes etárias, seguido do galho, folha e casca.

ABSTRACT The aim of this study was to determine the effect of age and plant spacing on biomass production of four forestry tree species: Acacia mearnsii De Wild, Eucalyptus grandis W. Hill ex Maiden, Mimosa scabrella Benth, and Ateleia glazioviana Baill. The following spacings of plants at the ages of 1, 3, and 5 years after planting were considered: 2.0 × 1.0 m, 2.0 × 1.5 m, 3.0 × 1.0 m, and 3.0 × 1.5 m. The study was installed in randomized complete block design. Biomass was determined by weighing different components of the trees after harvesting. Plant spacing affected biomass production of forestry trees at different ages after planting. Dense spacings produced larger quantities of biomass than less dense spacing. The tree species differed in biomass production: Eucalyptus grandis produced the largest quantity (325.1 t ha-1), followed by Acacia mearnsii (239.3 t ha-1), Mimosa scabrella (53.4 t ha-1), and Ateleia glazioviana (32.1 t ha-1). Wood biomass represented the biomass component with the largest production over time, which showed an increasing proportion throughout the age groups, followed branch, leaf, and bark biomass.

RESUMO Com o propósito de compreender e relacionar o desempenho fisiológico das plantas às adversidades meteorológicas, objetivou-se determinar a resposta fisiológica de plantas de cidró (Aloysia triphylla) submetidas a diferentes níveis de reposição hídrica no verão e inverno. O trabalho foi desenvolvido em ambiente protegido, no município de Frederico Westphalen-RS. O delineamento utilizado foi blocos completos ao acaso em um esquema fatorial 2x4x2, sendo duas estações do ano (verão e inverno), quatro disponibilidades hídricas (50, 75, 100, 125% da ETo) e duas faces da folha (abaxial e adaxial), com quatro repetições. Avaliou-se a resistência à difusão de vapor pelos estômatos (s/cm), transpiração (µg/cm2/s) e a temperatura da folha (°C), em ambas as faces da folha, no período compreendido entre 10 às 11, 13 às 14 e 16 às 17 horas, aos 45 dias após o início da estação. A resposta fisiológica das plantas de cidró foi influenciada pelo nível de reposição hídrica, de forma diferenciada entre as estações do ano. A estação do verão foi responsável pelos maiores valores de radiação fotossinteticamente ativa incidente, temperatura da folha, transpiração e resistência à difusão de vapor pelos estômatos. A redução na disponibilidade hídrica promoveu o aumento nos valores de resistência estomática ao fluxo de vapor, reduzindo a taxa de transpiração das folhas. A radiação fotossinteticamente ativa incidente, temperatura da folha e transpiração apresentaram maiores valores na face abaxial da espécie estudada.

ABSTRACT We aimed to determine the response of cidró (Aloysia triphylla) submitted to different water replacement levels in summer and winter, in order to understand and relate the plant physiological performance under adverse weather conditions. The research was developed in a protected environment, in the city of Frederico Westphalen, Rio Grande do Sul State, Brazil. The experimental design was completely randomized, arranged in a 2x4x2 factorial scheme, two seasons of the year (summer and winter), four water availability (50, 75, 100, 125% of ETo) and two faces of leaf (abaxial and adaxial), with four replications. There were evaluated stomatal vapor diffusion resistance (s/cm), transpiration (µg/cm2/s) and leaf temperature (°C), on both faces of the leaf from 10 a.m. to 11 a.m., 1 p.m. to 2 p.m. and 4 p.m. to 5 p.m., 45 days after the beginning of the season. We concluded that the physiological response of cidró plants was influenced by water replacement levels, in a different way according to the seasons of the year. The summer season was responsible for the highest values of incident photosynthetically active radiation, leaf temperature, transpiration and stomatal vapor diffusion resistance. The reduction in water availability promoted the increase in values for stomatal resistance to vapor flow, reducing leaf transpiration rate. The incident photosynthetically active radiation, leaf temperature and transpiration showed highest values on the abaxial face of the studied species.

RESUMO: A utilização da madeira para fins energéticos é elevada no país, motivada principalmente pela diversidade de espécies que apresentam potencialidade de geração de energia. Esse trabalho teve como objetivo determinar o efeito da idade e do espaçamento de plantio nas propriedades energéticas em diferentes compartimentos da biomassa: madeira, casca e galho da espécie Ateleia glazioviana Baill, distribuída em diferentes espaçamentos: 2,0x1,0m, 2,0x1,5m, 3,0x1,0m e 3,0x1,5m, no 1º, 3º e 5º ano após o plantio. O presente trabalho foi realizado em um experimento instalado em delineamento experimental de blocos completos casualizados em três repetições. Foram determinadas as variáveis biomassa (BIO), poder calorífico superior (PCS), massa específica básica (ME), produtividade energética (PE), densidade energética (DE), teor de carbono fixo (CF), teor de material volátil (MV) e teor de cinzas (CZ). Concluiu-se que as diferentes idades das árvores influenciam significativamente todas as variáveis analisadas, proporcionando uma distribuição crescente da BIO, PE, CF, ME e DE, sem tendência a estabilização ao longo do tempo. Os quatro espaçamentos de plantio induzem a diferentes produções de BIO e PE, com uma tendência de redução dos seus valores com o aumento do espaçamento, em todos os períodos avaliados. Os quatro compartimentos das árvores induzem a um efeito significativo das variáveis BIO, PE, PCS, CF, MV e CZ.

ABSTRACT: The use of wood for energy purposes is common in Brazil, motivated mainly by the diversity of species that present a potential for energy production. The objective of this study was to determine the effects of tree age and spacing on the energy properties of different components of the biomass: wood, bark, and branch of the species Ateleia glazioviana Baill, distributed at different spacings: 2.0 × 1.0m, 2.0 × 1.5m, 3.0 × 1.0m, and 3.0 × 1.5m in the 1st, 3rd, and 5th year after the planting. The experiment was carried out in a randomized complete block design with three replications. The following variables were determined: biomass (BIO), gross calorific power (GCV), basic density (BD), energy productivity (EP), energy density (ED), fixed carbon content (FCC), volatile material content (VMC), and ash content (AC). Different tree ages significantly affected all analyzed variables, increasing the distribution of BIO, EP, AC, BD, and ED with no tendency to stabilize over time. The four tree spacings induced different levels of BIO and EP, with a trend toward lower values with increasing tree spacing for all assessed periods. The four components of the trees had a significant effect on the variables BIO, EP, GCV, FCC, VMC, and AC.

RESUMO Objetivou-se, com este trabalho, avaliar o efeito de doses e de fontes de nitrogênio sobre as variáveis morfológicas relacionadas com a produtividade de dois cultivares de girassol. O delineamento utilizado foi o de blocos ao acaso, em esquema fatorial 5 x 2 x 2, com cinco doses de adubação nitrogenada (0, 40, 80, 120 e 160 kg ha-1), duas fontes de nitrogênio (N), fertilizante convencional (ureia), com 44% de N e fertilizante nitrogenado de liberação lenta, com 26% de N, e dois cultivares de girassol (Olisun 3 e Aguará 6), com três repetições. Avaliaram-se altura de plantas, índice de área foliar, interceptação de radiação global, massa de mil aquênios, diâmetro do capítulo, produtividade de aquênios e estimou-se o teor de clorofila. Doses crescentes de N promovem aumento do índice de área foliar e de interceptação da radiação global, do florescimento pleno e da produtividade do girassol, de forma diferenciada entre cultivares. O fertilizante nitrogenado de liberação lenta possibilita índice de área foliar, no início de enchimento de aquênios, maior que o obtido por uso de ureia, a partir da dose de 120 kg ha-1 de N. O cultivar Olisun 3 apresenta maior índice de área foliar no florescimento pleno e maior produtividade, a partir das doses de 80 e de 120 kg ha-1 de N, respectivamente.

ABSTRACT The objective of this study was to evaluate the effect of doses and sources of nitrogen on themorphological variables related to yield of two sunflower cultivars. The design was randomized blocks in a factorial design 5 x 2 x 2, with five doses of nitrogen (0, 40, 80, 120 and 160 kg ha-1), two sources of nitrogen (N), fertilizer conventional (urea), with 44% N, and the fertilizer nitrogenated of slow-release, with 26% N, and two sunflower cultivars (Olisun 3 and Aguará 6) with three replications. Was evaluated theplant height, leaf area index, global radiation interception, thousand achenes weight, diameter of achenes, yield and estimate the chlorophyll content. Increasing doses of N promotes an increase in the leaf area index and interception of global radiation at flowering full and the sunflower yield, in a differentiated way between cultivars. The use of fertilizer nitrogenated of slow-release provides higher leaf area index in the beginning of achenes filling, compared the use of urea, from the dose of 120 Kg ha-1 N. The cultivar Olisun 3 shows greater leaf area index at flowering full and yield, from the doses of 80 and 120 kg ha-1 N, respectively.