http://lattes.cnpq.br/2049175188318520; SILVA, Luderlândio de Andrade.
Resumo:
plants are of great socioeconomic importance for many countries and, in particular, for Brazil, due to their nutritional and medicinal properties. In regions with qualitative and quantitative limitations of water resources, however, such as the Brazilian semi-arid region, it is necessary to use tolerant canopy/rootstock combinations, which can improve the growth and productivity rates of the crop's fruits. In this context, the objective was to select salinity tolerant rootstocks for the acid lime tree 'Tahiti', based on ecophysiological and production parameters, applying the mixed model (REML/BLUP) and principal components (ACP) methodology. The combinations were formed by grafting the acid lime tree 'Tahiti' into ten rootstocks, obtaining the seedlings in the Embrapa Cassva & Frutis, in Cruz das Almas, BA. Plants were cultivated under salt stress in drainage lysimeters with a capacity of 150 dm3, in a randomized block design (three replications), factorial scheme 10 x 2, referring to ten citrus genotypes (nine triple hybrids and the 'Rangpur’ lime, as a comparison) and two levels of electrical conductivity of the irrigation water (CEa: S1=0.3 and S2=3.0 dS m-1). The plot consisted of a lysimeter, containing a plant. The beginning of irrigation with the two water qualities occurred 15 days after transplanting the seedlings, extending during the first two years of cultivation, with evaluation of the plants in relation to growth, physiological and production variables. The data obtained were evaluated by analysis of variance by the 'F' test. In cases of significance, the mean grouping test (Scott and Knott up to 5% probability) was performed for the rootstock factor during the seedling formation phase at each level of salinity of the water studied. using the REML/BLUP method (restricted maximum likelihood / best unbiased linear prediction) to identify the degree of tolerance to salt stress of the 'Tahiti'/rootstock combinations, accompanied by GT Biplot analysis, with the aid of the R software (R Development Core Team, 2014).