FertiliCalc: A Decision Support System for Fertilizer Management

Abstract

Rational fertilizer management is crucial in the efficient use of resources that are basically non-renewable and that can have a great environmental impact when used without scientific basis. The availability of scientifically sound decision-making tools for rational fertilization is scarce. We have developed a Windows program to calculate the required seasonal N, P and K rates, and the most cost-effective combination of commercial fertilizers. The tool also provides estimates of the Ca, Mg and S balances in the field resulting from the fertilizer program chosen. Novel aspects of the calculations include the development of stochastic flexible fertilizer programs for N and the calculation of acidification and N losses. Regarding P and K, estimations are provided on the grounds of threshold values of usual availability indexes, something frequently unknown by final users. Also, it allows the users to determine the best complex fertilizer for pre-plant applications to avoid blending of simple fertilizers at the farm, a task usually complex for farmers. The application may be useful both to the fertilizer supply and demand sides. In addition, it may be used for teaching as it helps understanding the rationale behind this management practice.

Appendices

Appendix A1. Calculation of N Requirements

N fertilizer requirements (N rate) are calculated from:

$$N rate = \frac{{N_{end} + \left( {1 + f_{NR} } \right)\left( {N_{yield} + N_{res} } \right) - k_{im} F_{res} N_{res}^{'} - f_{NR} \left( {N_{yield}^{'} + N_{res}^{'} } \right) - N_{other} }}{{\left( {1 - n} \right)}}$$

(6)

In this equation, N_end represents the final soil inorganic N (residual N). FertiliCalc uses a fixed value of 10 kg N ha⁻¹ assuming that crops are unable to recover N below that threshold. f_NR is the ratio of N in roots to N in shoots. N_yield and N_res refer to N accumulated in the harvest organ and residues of the present crop, respectively, while their homologous N′_yield and N′_res correspond to the previous crop in the rotation. These values are easily calculated from the product of concentrations of N in harvested organs and residues and their biomass (Quemada et al. 2016a). The coefficient k_im would have a maximum value of 1 if all the aboveground residues were mineralized with no loss. Lower values are expected if the residues are not incorporated by tillage or when the N concentration in residues is low. FertiliCalc adopts different values depending on whether the crop is a legume and whether it is tilt. F_res is the fraction of residues that are left in the field (user-defined input; otherwise the application provides default values depending on the crop). N_other is the total N received by atmospheric deposition, symbiotic fixation and irrigation water. In the case of non-legume crops, FertiliCalc adopts a default value. For legume crops, the application calculates N_other as a fraction of the crop N (f_fix):

$$N_{other} = f_{fix} \left( {1 + f_{NR} } \right)\left( {N_{yield} + N_{res} } \right)$$

(7)

where f_fix takes different values depending on the type of legume crop (annual or perennial) and the percentage of soil organic matter (Quemada et al. 2016a). Finally, the coefficient n in Eq. (6) represents the fraction of applied N that is lost (leaching, volatilization, denitrification). Depending on soil texture, FertiliCalc assumes that leaching ranges from 20% (sandy) to 2% (clayish). The rates of volatilization of ammonia and denitrification are determined according to Quemada et al. (2016b).

The model assumes that most of N supplied by mineralization, atmospheric deposition, symbiotic fixation and contained in the irrigation water, are taken up by crops with no losses. Table 4 provides a list with the values of the aforementioned parameter used by FertiliCalc in the calculation of N rate.

Table 4 List of parameter values adopted by FertiliCalc for the calculation of N requirements

Appendix A2. Conversion Factors for Soil P Tests

When soil P data available have not been determined by the Olsen method, FertiliCalc estimates the equivalent Olsen STL (STL_Olsen) as:

$$STL_{Olsen} = k\,STL_{i}$$

(8)

where STL_i is the soil test level determined by the method “i” and k a conversion factor that is method-specific. Values for k have been calculated from data reported in Neyroud and Lischer (2003) and are presented in Table 5.

Table 5 Values of the coefficient converting values of a given soil P test into its equivalent for the Olsen method