Proposing Nutrient Reference Values in Peach Leaves Based on Bayesian Segmented Quantile Regression

Abstract A. Purpose: The aim of this study is to propose new reference values for nutrients in leaves of peach trees grown in Southern Brazil by using Bayesian Segmented Quantile Regression (BSQR). b. Methods: A database containing yield data and nutrient concentrations in leaves of two peach tree cultivars, 'PS' (fresh market) and 'Maciel' (processing), was used. Critical levels (CL) and sufficiency ranges (SR) values were obtained through the BSQR method. c. Results: The CL and SR of N and K in peach leaves were the same in both cultivars (N = 31 g kg −1 , K = 23.5 g kg −1 , respectively). However, the CL and SR values in P, Ca, Mg, S, B, Cu and Fe were different between the cultivars used in the study. The study also enabled establishing reference values for S in peach leaves (CL = 0.85 g kg −1 and SR = 0.70–1.00 g kg −1 ). These values are scarce in the literature and do not exist in the official fertilizer recommendation manual for RS and SC states, Southern Brazil. d. Conclusions: The BSQR method proved to be efficient in obtaining CL and SR from a peach orchard database. The results proved that CL and SR best adjusted to the regions and cultivars in comparison to generalist SR. This information will enable more rational use of fertilizers.