ABSTRACT: Leaf hyperspectral reflectance has been used to estimate nutrient concentrations in plants in narrow bands of the electromagnetic spectrum. The aim of this study was to estimate leaf nutrient concentrations using leaf hyperspectral reflectance and verify the variable selection methods using the partial least squares regression (PLSR). Two studies were carried out using stands with Eucalyptus clones. Study I was established in Eucalyptus stands with three clones, classifying leaves into five colour patterns using the Munsell chart for plant tissues. Immediately after leaf collection, leaf reflectance was read and the chemical analysis was performed. Study II was carried out in commercial clonal stands of Eucalyptus performing the same leaf sampling and chemical analysis as used in Study I. All leaf reflectance spectra were smoothed and three more pre-processing procedures were applied. In addition, three methods of PLSR were tested. The first derivative was more accurate for predicting nitrogen ( R c v 2 = 0.95), phosphorous ( R c v 2 = 0.93), and sulphur concentration ( R c v 2 = 0.85). The estimates for concentrations of calcium ( R c v 2 = 0.81), magnesium ( R c v 2 = 0.22), and potassium ( R c v 2 = 0.76) were more accurate using the logarithm transformation. Only the estimates for iron concentrations were performed with higher accuracy ( R c v 2 = 0.35) using the smoothed reflectance. The copper concentrations were more accurate ( R c v 2 = 0.78) using the logarithm transformation. Concentrations of boron ( R c v 2 = 0.68) and manganese ( R c v 2 = 0.79) were more accurate using the first derivative, while zinc ( R c v 2 = 0.31) concentration was most accurate using the second derivative.

ABSTRACT Determination of leaf nutrient concentrations is traditionally performed by carrying out destructive procedures, requiring laboratory chemical analysis, specialized equipment, and skilled labor. However, technological advances in recent years have now made it possible to analyze leaf nutrient concentrations using non-destructive methods such as leaf reflectance. This study evaluates the relationship between leaf reflectance and nutrient concentration in order to develop and apply nutrient indices by proximal sensing data in plantations of Eucalyptus. Two experiments were carried out for the development, application and validation of nutrient indices in Eucalyptus stands in the municipalities of Lassance and Três Marias, Minas Gerais state, Brazil. Study I was undertaken in Eucalyptus stands with 25 months old and, three clones. Leaves from the lower crown were visually classified into five color patterns using the Munsell chart for plant tissues. Study II was carried out in commercial stands of a hybrid of Eucalyptus urophylla ST Blake, whith 9, 12, 15 and 25 months old. Analysis revealed strong relationships between leaf nutrients and leaf reflectance in the visible and near infrared regions (400 - 900 nm) of the light spectrum. Correlation analysis between leaf reflectance and nutrients can be useful in an exploratory analysis of leaf nutrient concentration. Some nutrient indices developed (NI, PI, SI and CuI) provided satisfactory estimates of leaf nutrient concentration in Eucalyptus stands. The nutrient indices developed in this study may be a useful alternative to laboratory chemical analysis.

ABSTRACT Nitrogen (N) has commonly been applied in Eucalyptus stands in Brazil and it has a direct relation with biomass production and chlorophyll content. Foliar N concentrations are used to diagnose soil and plant fertility levels and to develop N fertilizer application rates. Normally, foliar N is obtained using destructive methods, but indirect analyses using Vegetation Indexes (VIs) may be possible. The aim of this work was to evaluate VIs to estimate foliar N concentration in three Eucalyptus clones. Lower crown leaves of three clonal Eucalyptus plantations (25 months old) were classified into five color patterns using the Munsell Plant Tissue Color Chart. For each color, N concentration was determined by the Kjeldahl method and foliar reflectance was measured using a CI-710 Miniature Leaf Spectrometer. Foliar reflectance data were used to obtain the VIs and the VIs were used to estimate N concentrations. In the visible region, the relationship between N concentration and reflectance percentage was negative. The highest correlations between VIs and N concentrations were obtained by the Inflection Point Position (IPP, r = 0.97), Normalized Difference Red-Edge (reNDVI, r = 0.97) and Modified Red-Edge Normalized Difference Vegetation Index (mNDI, r = 0.97). Vegetation indexes on the red edge region provided the most accurate estimates of foliar N concentration. The reNDVI index provided the best N concentration estimates in leaves of different colors of Eucalyptus urophylla × grandis and Eucalyptus urophylla × urophylla (R2 = 0.97 and RMSE = 0.91 g kg−1).

O gênero Hymenaea pertence à família Leguminosae, subfamília Caesalpinoideae, com nome popular de jatobá. No gênero Hymenaea, foram descritas aproximadamente 25 espécies. Já no Brasil, verifica-se a presença de 13 espécies, com destaque para a Hymenaea martiana Hayne e a Hymenaea courbaril Linneaus. Na literatura, há vários trabalhos que tratam principalmente da quebra de dormência das sementes e da produção de mudas, sobretudo da espécie Hymenaea courbaril Linneaus, porém essas avaliações por matriz ainda são incipientes. Este trabalho teve por objetivo avaliar e comparar a germinação e o crescimento inicial de mudas entre as espécies e as matrizes de Hymenaea martiana Hayne e Hymenaea courbaril Linneaus. Foram desenvolvidos dois experimentos: o primeiro avaliou a germinação, e o segundo, o crescimento inicial de mudas. As duas espécies avaliadas apresentaram comportamentos distintos entre a germinação e o crescimento inicial das mudas, sendo a Hymenaea courbaril Linneaus a espécie com maior média para ambas variáveis, porém não apresentou variabilidade fenotípica entre matrizes.

The Hymenaea genus belongs to the family Leguminosae, subfamily Caesalpinoideae with the common name jatobá. Approximately 25 species have been described in the genus Hymenaea and 13 of them are present in Brazil. Among them we can highlight the species Hymenaea martiana Hayne and Hymenaea courbaril Linneaus. In literature, there are several studies dealing mainly the dormancy breakage and the seedling production, mainly with Hymenaea coubaril Linneaus, but these reviews per matrices are incipient. This study aimed to evaluate and compare the germination and seedling growth between species and between matrices of the Hymenaea martiana Hayne and Hymenaea courbaril Linneaus. Two experiments were conducted: the first evaluated the germination and the second the seedlings growth. The two species studied showed different behavior between germination and the seedling growth, with Hymenaea courbaril Linneaus showing the greatest average for both variables. However, this did not show phenotypic variability between matrices.