Gledson Luiz Pontes de Almeida

Poultry production faces increasing challenges that directly affect animal welfare, productivity, and the sustainability of production systems. This review aimed to map the international scientific literature on housing systems for broilers and laying hens from 2014 to 2024, identifying research trends, knowledge gaps, and the main topics investigated in recent years. A total of 141 articles indexed in the Scopus and Web of Science databases were analyzed using the bibliometrix package in RStudio version 2026.08.2+200. The results revealed a significant increase in the number of publications from 2020 onward, with the United States, Brazil, and Germany emerging as the leading contributing countries. Research focused on topics such as heat stress, animal welfare, environmental control, building materials, and biosecurity. The analysis of scientific collaboration networks revealed a predominance of international publications and limited international collaboration, particularly in Brazil. Overall, the findings highlight the growing need for interdisciplinary and technology-oriented research aimed at increasing poultry productivity, thereby promoting sustainability, animal welfare, and production efficiency.

Observing pig behavior, such as feed intake, water intake, and resting behavior, is essential for improving the well-being of these animals. However, monitoring such behaviors by traditional methods can be exhausting for both humans and animals, interfering with their development. The research aimed to identify behavioral patterns of pigs in an air-conditioned environment through computer vision. Microcameras were installed in the animals’ stalls to generate videos over an experimental period of 92 days and the temperature and humidity of the air were simultaneously recorded. The physiological variables of the animals were collected to identify whether they were under heat stress. To recognize the drinking, eating, standing and lying behavior of pigs, YOLOv5 was trained and then the model was used to detect the animals. Regions in the images corresponding to the feeders and drinkers were established. To identify feeding behavior and water intake, criteria based on the occupation of the feeding zone by pigs detected in the standing position were established. The results showed that the trained model achieved an average accuracy rate of 97.3% and an average recall of 96.1% in animal detection. The model exhibited 97.5% accuracy and 97.0% recall rates in recognizing the feeding behavior and water consumption of pigs. The proposed method can be used in videos or images and minimizes the need for manual intervention, offering an efficient means of monitoring pig behavior in agricultural environments and contributing to the productivity of pig farming operations.

The present study aimed to apply machine learning algorithms to estimate respiratory rate (RR, breaths min⁻¹) and rectal temperature (RT, °C) as indicators of thermal stress in Dorper breeding rams, based on environmental and thermal variables obtained through infrared thermography. The algorithms Random Forest (RF) and Support Vector Regression (SVR) with radial kernel were employed, using ocular globe temperature (OGT), air temperature (AT), relative humidity (RH), and coat surface temperature (CST) as predictor variables, and rectal temperature (RT) and respiratory rate (RR) as response variables. Data were collected on a property located in Garanhuns, Pernambuco State, Brazil, under two environmental conditions (with and without climate control), totaling 20 monitored animals and 120 paired observations. Model performance was evaluated using the coefficient of determination (R²), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE), complemented by cross-validation (k-fold = 10), and model interpretability was assessed using SHapley Additive exPlanations (SHAP) to quantify the contribution of each predictor variable to model predictions. The results indicated that the RF model showed superior performance in predicting the physiological variables RR and RT, with higher coefficients (RR: R² = 0.858; RT: R² = 0.687) and lower error values. For RR, the RF model achieved RMSE = 16.38 and MAE = 13.33; while for RT, the errors were RMSE = 0.217 and MAE = 0.154. In contrast, the radial kernel SVR model showed lower performance, with R² values of 0.742 (RR) and 0.533 (RT), and RMSE and MAE values of 21.05 and 17.38 for RR, and 0.262 and 0.196 for RT, respectively. The application of machine learning-based models proved to be a viable and accurate alternative for estimating physiological indicators of thermal stress, contributing to the development of automated thermal management strategies for sheep in the Brazilian semi-arid region. The proposed data-driven approach demonstrates that low-cost thermal sensors combined with explainable artificial intelligence can support automatic decision-making for climate adaptation and animal welfare in semi-arid sheep production systems.

Estimated population growth and increased demand for food production bring with them the evident need for more efficient and sustainable production systems. Because of this, computer vision plays a fundamental role in the development and application of solutions that help producers with the issues that limit livestock production in Brazil and the world. In addition to being stressful for the producer and the animal, the conventional pig weighing system causes productive losses and can compromise meat quality, being considered a practice that does not value animal welfare. The objective was to develop a computational procedure to predict the live weight of pigs in the growth and finishing phases, through the volume of the animals extracted through the processing of 3D images, as well as to analyze the real and estimated biometric measurements to define the relationships of these with live weight and volume obtained. The study was conducted at Roçadinho farm, in the municipality of Capoeiras, located in the Agreste region of the state of Pernambuco, Brazil. The variables weight and 3D images were obtained using a Kinect^®—V2 camera and biometric measurements of 20 animals in the growth phase and 24 animals in the finishing phase, males and females, from the crossing of Pietrain and Large White, totaling 44 animals. To analyze the images, a program developed in Python (PyCharm Community Edition 2020.1.4) was used, to relate the variables, principal component analyses and regression analyzes were performed. The coefficient of linear determination between weight and volume was 73.3, 74.1, and 97.3% for pigs in the growing, finishing, and global phases, showing that this relationship is positive and satisfactorily expressed the weight of the animals. The relationship between the real and estimated biometric variables had a more expressive coefficient of determination in the global phase, having presented values between 77 and 94%.

Forage scarcity, intensified by climate variability and edaphoclimatic limitations in the Brazilian semiarid region, challenges regional livestock production. In this context, forage palm is a strategic alternative due to its drought resistance and environmental adaptability. However, little is known about the spatial and temporal dynamics of its cultivation. This study aimed to characterize the spatio-temporal dynamics of forage palm cultivation in Capoeiras-PE between 2019 and 2022 using remote sensing data and multitemporal analysis of the Normalized Difference Vegetation Index (NDVI), processed via Google Earth Engine. Experimental areas with Opuntia stricta (“Mexican Elephant Ear”) and Nopalea cochenillifera (“Miúda”) were monitored, with field validation and descriptive statistical analysis. NDVI values ranged from −0.27 to 0.93, influenced by rainfall, cultivar morphology, and seasonal conditions. The “Miúda” cultivar showed a lower coefficient of variation (CV%), indicating greater spectral stability, while “Orelha de Elefante Mexicana” was more sensitive to climate and management, showing a higher CV%. Land use and land cover (LULC) analysis indicated increased sparse vegetation and exposed soil, suggesting intensified anthropogenic activity in the Caatinga biome. Reclassified NDVI enabled spatial estimation of forage palm, despite sensor resolution and spectral similarity with other vegetation. The integrated use of satellite data, field validation, and geoprocessing tools proved effective for agricultural monitoring and territorial planning.

The state of Pernambuco is the second largest milk producer in the Northeast. The region has great temporal and spatial variability in rainfall, which directly influences agriculture and livestock farming. The aim of this study was to analyse the interaction between meteorological elements, the Normalized Difference Vegetation Index (NDVI), agricultural production and milk production using multivariate principal component analysis, considering a 20-year historical series in the municipality of Capoeiras-PE. The years with the highest milk production in Capoeiras occurred between 2017 and 2021. NDVI proved to be associated with increased productivity for all crops, as well as being correlated with rainfall and relative humidity. During the period of growth in dairy production, a strong relationship was observed with the average annual temperature, possibly due to the adoption of new management techniques focused on animal welfare.In contrast to the influence of temperature, the results also indicated a strong correlation between NDVI and rainfall over the period studied, demonstrating the efficiency of this index in identifying the best harvests for the crops analyzed. It can be concluded that NDVI makes it possible to establish relationships between geoprocessed data and meteorological parameters, exerting a direct or inversely proportional influence on these factors.

A região semiárida possui uma economia diversificada no setor agropecuário, no entanto, a seca gera grande prejuízo a atividade agrícola. Assim, a palma forrageira, surge como uma alternativa para o suprimento da alimentação do gado leiteiro. Com essa pesquisa, determinou-se a assinatura espectral da palma forrageira em diferentes idades de cultivos, através do sensoriamento remoto e geoprocessamento, utilizando imagens de satélite. A análise multivariada mostrou que nos anos com precipitação acima da média está em oposição ao agrupamento que contém os menores valores das reflectâncias (nas bandas do visível e do infravermelho próximo), que indicam os períodos de maior cobertura da superfície do solo pela palma forrageira em relação a vegetação nativa. Contudo, foi possível o monitoramento da palma forrageira com diferentes graus de desenvolvimento, subsidiando as tomadas de decisão, no que concerne o manejo das áreas cultivadas com palma para aumento de produção leiteira no semiárido.

CHARACTERIZATION, CHEMOMETRIC ANALYSIS, AND PRELIMINARY ADSORPTION TESTS OF TEXTILE DYES ON BACTERIAL CELLULOSE: A TECHNOLOGY FOR EFFLUENT TREATMENT WITH A PROPOSAL FOR THE DEVELOPMENT OF A SUSTAINABLE ECONOMY. To mitigate the environmental impacts of dyes in the textile industry, this research sought to investigate the removal potential of a bacterial cellulose (BC) produced by Komagataeibacter xylinus. The BC was produced under controlled conditions and then the physicochemical, structural, and morphological properties of the biosorbent were characterized using Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and thermogravimetry (TG). Ten dye samples were collected from the textile laundries in Pernambuco to carry out the adsorption studies. The chemical composition of the dyes was assessed by electronic absorption spectroscopy (UV-Vis), inductively coupled plasma optical emission spectroscopy (ICP-OES), and TG. They were also statistically analyzed using principal component analysis (PCA). The treatment tests were conducted in a 24-h adsorption system using ∼ 100 mg of BC and 20.0 mg L-1 of dyes. The most satisfactory response was for phthalocyanine class dyes, with qe (amount of analyte adsorbed at equilibrium) of 1.7234 mg g-1 under acid conditions. In the thermal analysis, it was defined that degradation of CB at 555 °C allowed recovery of approximately 65% of C10 dye (reactive blue 21) for the impregnated material. A hypothesis was also formulated as an approach to treat textile water waste, to produce an "eco-dye".

Mastitis is a disease that is considered an obstacle in dairy farming. Some methods of diagnosing mastitis have been used effectively over the years, but with an associated relative cost that reduces the producer’s profit. In this context, this sector needs tools that offer an early, safe, and non-invasive diagnosis and that direct the producer to apply resources to confirm the clinical picture, minimizing the cost of monitoring the herd. The objective of this study was to develop a predictive methodology based on sequential knowledge transfer for the automatic detection of bovine subclinical mastitis using computer vision. The image bank used in this research consisted of 165 images, each with a resolution of 360 × 360 pixels, sourced from a database of 55 animals diagnosed with subclinical mastitis, all of which were not exhibiting clinical symptoms at the time of imaging. The images utilized in the sequential learning transfer were those of MammoTherm, which is used for the detection of breast cancer in women. The optimized model demonstrated the most optimal network performance, achieving 92.1% accuracy, in comparison to the model with manual search (86.1%). The proposed predictive methodologies, based on knowledge transfer, were effective in accurately classifying the images. This significantly enhanced the automatic detection of both healthy animals and those diagnosed with subclinical mastitis using thermal images of the udders of dairy cows.