Abstract A novel pigment of β-LiAlSi2O6:Cr ceramic was developed by a partial proteic sol-gel process, using gelatin as a ligand. X-ray diffraction (XRD), energy dispersive spectroscopy (EDS) and visible and near-infrared diffuse reflectance spectroscopy (vis-NIR DRS) were performed. XRD results confirmed that the crystal structure of the lattice corresponded to β-LiAlSi2O6 (or β‑spodumene), and the addition Cr3+ ions by doping did not interfere in the formation of this crystalline phase. EDS confirmed the homogeneous existence of Cr3+ dopant in β‑LiAlSi2O6 particles. From the vis‑NIR DRS, selective absorption of visible light wavelengths was identified in the bands of 425 nm and 600 nm, resulting in the perception of a yellowish-green color when β-LiAlSi2O6 is doped with Cr3+. CIE-XYZ colorimetric coordinates were generated to characterize the resulting colors. The obtained results demonstrated the viability of β‑LiAlSi2O6:Cr synthesis by a proteic sol‑gel route and its great potential for obtaining a yellowish-green ceramic pigment.

ABSTRACT This paper aimed to develop predictive models to determine total soluble solids, firmness, and ripening stages of ‘Pacovan’ bananas, using Vis-NIR spectroscopy and machine learning algorithms. A total of 384 bananas were divided into different days of storage (0, 3, 6, 9, 12, 15, 18, and 21 days) at two temperatures (25°C and 20°C). Bananas were subjected to spectral analysis using a spectrometer operating in spectral range of 350 – 2500 nm. Physicochemical parameters of quality, total soluble solids, and firmness were determined by reference analyses. Different machine learning algorithms were used to develop regression models and supervised classification. The best model for total soluble solids was the Random Forest with variable selection, showing an R2cv of 0.90 and RMSECV of 2.31. The best model for firmness was the Support Vector Machine with variable selection, showing an R2cv of 0.84 and RMSECV of 7.98. The best classification model for different ripening stages was the Multilayer Perceptron with variable selection, which achieved the precision of 74.22%. Therefore, Vis-NIR spectroscopy associated with machine learning algorithms is a promising tool for monitoring the quality and ripening stages of ‘Pacovan’ bananas.

Objetivou-se com o trabalho avaliar o desempenho ponderal de dois genótipos de ovinos, Santa Inês e F1 Dorper x Santa Inês em pastagem de Panicum maximum cv. Massai. Foram utilizados oito fêmeas e oito machos, de cada genótipo, em delineamento inteiramente casualizado, em esquema fatorial 2x2x2 (dois genótipos, dois sexos e duas idades de abate). Os dados foram submetidos à análise de variância com aplicação do teste de Scot-Knott com probabilidade de 5%. Avaliaram-se peso ao nascimento, peso a desmama, peso final, ganho de peso médio diário e ganho de peso total. Houve efeito significativo do genótipo para peso ao nascimento e peso a desmama com os cordeiros Dorper x Santa Inês superiores. Para peso final não houve diferença significativa, pois o ganho de peso do desmame ao abate foi superior para o genótipo Santa Inês. Para ganho de peso total o efeito do genótipo foi importante visto que o F1 Dorper x Santa Inês foi superior ao Santa Inês, enquanto o ganho de peso médio final foi igual para os dois genótipos avaliados. O efeito de sexo não foi significativo para nenhuma das características avaliadas. Não houve diferença entre peso final, ganho de peso médio diário e ganho de peso total nas diferentes idades de abate. A utilização da raça Dorper não refletiu maior retorno quando se mantém seus produtos em sistemas exclusivamente a pasto. Há necessidade de estudos bioeconômicos para verificar a aplicabilidade desta prática na região Norte de Minas Gerais.

This work was carried out in order to evaluate the daily weight gain of two sheep genotypes, Santa Inês and F1 Dorper x Santa Inês. Eight females and eight males from each genotype, were used, in a entirely randomized design, in a factorial scheme 2 x 2 x 2 (two genotypes, two sexes and two slaughter ages). Data were submitted to analysis of variance with application of the Scott-Knott test at 5% probability. The birth weight, wean weight, final weight, average daily weight gain and total weight gain were evaluated. There was significant effect of genotype for birth weight and wean weight with superior Dorper x Santa Inês lambs. There was no significant difference for final weight, since weight gain from wean to slaughter was superior for Santa Inês genotype. For total weight gain, the effect of the genotype was important because F1 Dorper x Santa Inês was superior to, while daily weight gain was equal for the two evaluated genotypes. Sex effect was not significant for any evaluated characteristic. There was no difference between final weight, average daily weight gain and total weight gain for different slaughter ages. The use of Dorper breed did not reflect greater return when is intended to keep its products in systems exclusively for pasture. Bioeconomic studies are necessary for verifying applicability of that practice in the North of Minas Gerais.