Meta-analysis in the Selection of Groups in Varieties of Citrus

Brazil is the largest producer of oranges (Citrus sinensis) in the world. The nutrient management of tree orchards is designed from experiments with a limited number of varieties. This knowledge is transferred to other varieties by diagnosing tissue nutrient composition and tree demand. Compositional data analysis has been first applied to tissue analysis of agricultural crops using centered log ratios with compositional nutrient diagnosis (CND-clr). The isometric log ratio (ilr) transformation is a new approach based on binary nutrient ratios and the principle of orthogonality (CND-ilr). We analyzed eleven nutrients: nitrogen (N), sulfur (S), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), boron (B), copper (Cu), zinc (Zn), manganese (Mn), and iron (Fe) in leaf tissue samples across 108 commercial plots (thirty-one grow Valencia, twenty-two Hamlim, twenty Pera, and thirty-five Natal). Nutrients were partitioned between macro- and micronutrients as well as anionic and cationic species. The effect size of varieties over Valencia was quantified by the mean and standard deviation of ilr values across ilr coordinates. Specific varietal nutrient profiles and ilr norms were defined. The nutrient profile of orange varieties could be classified into homogeneous groups to take advantage of fertilizer trials conducted on varieties of the same group. The Aitchison distance and a perturbation vector could be instrumental for diagnostic purposes and nutrient management.