Machine learning in the classification of asian rust severity in soybean using hyperspectral sensor.

Traditional monitoring of asian soybean rust severity is a time- and labor-intensive task, as it requires visual assessments by skilled professionals in the field. Thus, the use of remote sensing and machine learning (ML) techniques in data processing has emerged as an approach that can increase efficiency in disease monitoring, enabling faster, more accurate and time- and labor-saving evaluations. The aims of the study were: (i) to identify the spectral signature of different levels of Asian soybean rust severity; (ii) to identify the most accurate machine learning algorithm for classifying disease severity levels; (iii) which spectral input provides the highest classification accuracy for the algorithms; (iv) to determine a sample size of leaves that guarantees the best accuracy for the algorithms. A field experiment was carried out in the 2022/2023 harvest in a randomized block design with a 6x3 factorial scheme (ML algorithms x severity levels) and four replications. Disease severity levels assessed were: healthy leaves, 25 % severity, and 50 % severity. Leaf hyperspectral analysis was carried out over a wide range from 350 to 2500 nm. From this analysis, 28 spectral bands were extracted, seeking to distinguish the spectral signature for each severity level with the least input dataset. Data was subjected to machine learning analysis using Artificial Neural Network (ANN), REPTree (DT) and J48 decision trees, Random Forest (RF), and Support Vector Machine (SVM) algorithms, as well as a traditional classification method (Logistic Regression - LR). Two different input datasets were tested for each algorithm: the full spectrum (ALL) provided by the sensor and the 28 spectral bands (SB). Tests with different sample sizes were also conducted to investigate the algorithms' ability to detect severity levels with a reduced sample size. Our findings indicate differences between the spectral curves for the severity levels assessed, which makes it possible to differentiate between healthy plants with low and high severity using hyperspectral sensing. SVM was the most accurate algorithm for classifying severity levels by using all the spectral information as input. This algorithm also provided high classification accuracy when using smaller leaf samples. This study reveals that hyperspectral sensing and the use of ML algorithms provide an accurate classification of different levels of Asian rust severity, and can be powerful tools for a more efficient disease monitoring process.

High-throughput phenotyping using VIS/NIR spectroscopy in the classification of soybean genotypes for grain yield and industrial traits.

Employing visible and near infrared sensors in high-throughput phenotyping provides insight into the relationship between the spectral characteristics of the leaf and the content of grain properties, helping soybean breeders to direct their program towards improving grain traits according to researchers' interests. Our research hypothesis is that the leaf reflectance of soybean genotypes can be directly related to industrial grain traits such as protein and fiber contents. Thus, the objectives of the study were: (i) to classify soybean genotypes according to the grain yield and industrial traits; (ii) to identify the algorithm(s) with the highest accuracy for classifying genotypes using leaf reflectance as model input; (iii) to identify the best input data for the algorithms to improve their performance. A field experiment was carried out in randomized block design with three replications and 32 soybean genotypes. At 60 days after emergence, spectral analysis was carried out on three leaf samples from each plot. A hyperspectral sensor was used to capture reflectance between the wavelengths from 450 to 824 nm. Representative spectral bands were selected and grouped into means. After harvest, grain yield was assessed and laboratory analyses of industrial traits were carried out. Spectral, industrial traits and yield data were subjected to statistical analysis. Data were analyzed by the following machine learning algorithms: J48 (J48) and REPTree (DT) decision trees, Random Forest (RF), Artificial Neural Networks (ANN), Support Vector Machine (SVM), and conventional Logistic Regression (LR) analysis. The clusters formed were used as the output of the models, while two groups of input data were used for the input of the models: the spectral variables (WL) noise-free obtained by the sensor (450-828 nm) and the spectral means of the selected bands (SB) (450.0-720.6 nm). Soybean genotypes were grouped according to their grain yield and industrial traits, in which the SVM and J48 algorithms performed better at classifying them. Using the spectral bands selected in the study improved the classification accuracy of the algorithms.

Dynamics of major environmental disasters involving fire in the Brazilian Pantanal.

The 2020 environmental catastrophe in Pantanal has highlighted the fragility of environmental policies and practices for managing and fighting fires in this biome. Therefore, it is essential to know the causes and circumstances that potentiate these fires. This study aimed to: (I) assess the relationship between fire foci and carbon absorption (GPP), precipitation, and carbon dioxide (CO2) flux; (ii) analyze vegetation recovery using the differenced normalized burn ratio (ΔNBR) in Brazilian Pantanal between 2001 and 2022; and (iii) identify priority areas, where the highest intensities of fire foci have occurred, in order to guide public policies in Brazil to maintain local conservation. To this purpose, fire foci were detected using data from the MODIS MOD14/MYD14 algorithm, annual precipitation with CHIRPS (Climate Hazards Group InfraRed Precipitation with Station data), and CO2 flux using the MODIS/MODO9A1 product, and Gross Primary Production (GPP) with the MODIS/MOD17A2 product. The severity of the burned area was also assessed using the ΔNBR index and the risk areas were determined using the averages of these images. During the time series studied, a total of 300,127 fire foci were detected throughout the Pantanal, where 2020 had the highest number of foci and the lowest accumulated precipitation. The years with the highest precipitation were 2014 and 2018. The year 2018 was also the second year with the highest GPP value. The Pettit test showed a trend for 2008 and 2011 as the points of change in the CO2 flux and GPP variables. Principal component analysis clustered fire foci and precipitation on opposite sides, as well as GPP and CO2 flux, while ΔNBR clustered HS, MHS and MLS classes with the years 2020, 2019, 2002 and 2021. There was a high negative correlation between fire foci × rainfall and GPP × CO2 flux. The years with the largest areas of High severity (HS), Moderate-high severity (MHS) and Moderate-low severity (MLS) classes were 2020 and 2019, respectively. The most vulnerable areas for severe fires were the municipalities of Cáceres, Poconé, and Corumbá. The major fire catastrophe in 2020 is correlated with the low precipitation in 2019, the high precipitation in 2018, and the increased GPP, as well government policies unfavorable to the environment.

Understanding the combining ability of nutritional, agronomic and industrial traits in soybean F2 progenies.

Obtaining soybean genotypes that combine better nutrient uptake, higher oil and protein levels in the grains, and high grain yield is one of the major challenges for current breeding programs. To avoid the development of unpromising populations, selecting parents for crossbreeding is a crucial step in the breeding pipeline. Therefore, our objective was to estimate the combining ability of soybean cultivars based on the F2 generation, aiming to identify superior segregating parents and populations for agronomic, nutritional and industrial traits. Field experiments were carried out in two locations in the 2020/2021 crop season. Leaf contents of the following nutrients were evaluated: phosphorus, potassium, calcium, magnesium, sulfur, copper, iron, manganese, and zinc. Agronomic traits assessed were days to maturity (DM) and grain yield (GY), while the industrial traits protein, oil, fiber and ash contents were also measured in the populations studied. There was a significant genotype × environment (G × A) interaction for all nutritional traits, except for P content, DM and all industrial traits. The parent G3 and the segregating populations P20 and P27 can be used aiming to obtain higher nutritional efficiency in new soybean cultivars. The segregating populations P11 and P26 show higher potential for selecting soybean genotypes that combine earliness and higher grain yield. The parent G5 and segregant population P6 are promising for selection seeking improvement of industrial traits in soybean.

High-throughput phenotyping allows the selection of soybean genotypes for earliness and high grain yield.

BACKGROUND: Precision agriculture techniques are widely used to optimize fertilizer and soil applications. Furthermore, these techniques could also be combined with new statistical tools to assist in phenotyping in breeding programs. In this study, the research hypothesis was that soybean cultivars show phenotypic differences concerning wavelength and vegetation index measurements. RESULTS: In this research, we associate variables obtained via high-throughput phenotyping with the grain yield and cycle of soybean genotypes. The experiment was carried out during the 2018/2019 and 2019/2020 crop seasons, under a randomized block design with four replications. The evaluated soybean genotypes included 7067, 7110, 7739, 8372, Bonus, Desafio, Maracai, Foco, Pop, and Soyouro. The phenotypic traits evaluated were: first pod height (FPH), plant height (PH), number of branches (NB), stem diameter (SD), days to maturity (DM), and grain yield (YIE). The spectral variables evaluated were wavelengths and vegetation indices (NDVI, SAVI, GNDVI, NDRE, SCCCI, EVI, and MSAVI). The genotypes Maracai and Foco showed the highest grain yields throughout the crop seasons, in addition to belonging to the groups with the highest means for all VIs. YIE was positively correlated with the NDVI and certain wavelengths (735 and 790 nm), indicating that genotypes with higher values for these spectral variables are more productive. By path analyses, GNDVI and NDRE had the highest direct effects on the dependent variable DM, while NDVI had a higher direct effect on YIE. CONCLUSIONS: Our findings revealed that early and productive genotypes can be selected based on vegetation indices and wavelengths. Soybean genotypes with a high grain yield have higher means for NDVI and certain wavelengths (735 and 790 nm). Early genotypes have higher means for NDRE and GNDVI. These results reinforce the importance of high-throughput phenotyping as an essential tool in soybean breeding programs.