Comparison of Mehlich-3 and DTPA Soil Tests for Analysis of Comparison of Mehlich-3 and DTPA Soil Tests for Analysis of Micronutrients in Kansas Soils Micronutrients in Kansas Soils

Summary Mehlich-3 (M3) was designed as a multi-nutrient soil test procedure and has become common at soil testing labs across the U.S. In Kansas, Mehlich-3 is predominately used as a soil test for phosphorus (P) and potassium (K), but recent studies have also investigated the use of M3 for the extraction of base cations and cation exchange capacity estimation. However, data relating M3 to traditional methods for soil micronutrient extraction remain scarce. The objective of this study was to investigate the relationship between M3 and diethylenetriamine pentaacetate (DTPA) extractable copper (Cu), iron (Fe), manganese (Mn), and zinc (Zn) across a wide range of Kansas soils. Strong positive correlations were observed between M3 and DTPA for each metal (Fe, r = .91; Zn, r = .98; Cu, r = .92). Correlations between M3 and DTPA were positive but weak for Mn (r = 0.17). Regression analyses suggest these relationships were not one-to-one and were dependent on soil pH. Results from this study show that conversion of M3 to a “DTPA equivalent” is possible but should take soil pH into consideration, especially for Fe and Mn.


Introduction
The Mehlich-3 soil test procedure has become part of the routine workflow for soil analysis at soil testing labs across the U.S. This procedure allows for the simultaneous measurement of numerous essential plant nutrients from a single extraction and reduces lab operating costs and the cost of soil testing for farmers and homeowners (Rutter et al., 2022). The interpretation of a soil test requires knowledge of its relationship to nutrient uptake by crops, or its correlation to existing soil testing methods. However, data relating M3 extractable micronutrient metals to the conventional soil test (DTPA) for micronutrient metals are scarce for Kansas soils. Previous research conducted in other regions have included the investigation of relationships between these methods (Cancela et al., 2007;Iatrou et al., 2015). Iatrou et al. (2015) found strong correlations between M3 and DTPA for Zn and Cu in Greek soils, but a poor correlation for Mn. Positive correlations between M3 and DTPA extractable Fe were also observed, but the relationship was substantially influenced by soil pH. Similar observations were made by Cancela et al. (2007)

Procedures
Soil samples were selected randomly from those submitted to the lab by Kansas farmers and homeowners over a six-month period (a total of 308 soil samples for this study), and covered a wide range of pH, organic matter, and micronutrient content (Table 1). Soils were dried in a forced-air oven at 104°F (40°C) and ground to pass a 2-mm sieve using a flail-type soil mill (Custom Laboratory Equipment, MO, U.S.). Samples were stored at 70°F (21°C) until analysis. A summary of general soil characteristics can be found in Table 1.
The extraction procedures employed during this study are described in Chapter 9 of the Recommended Soil Testing Procedures for the North Central Region handbook (Whitney, 2015). Briefly, 10 g of soil were extracted with 20 mL of DTPA extracting solution (0.005 M DTPA, 0.01 M triethanolamine (TEA), pH = 7.3) on a reciprocating shaker at 180 rpm for 2 hours. Subsequent extracts were filtered using Ahlstrom 74 filter paper, and the concentrations of Cu, Fe, Mn, and Zn were measured using inductively coupled plasma atomic emission spectroscopy (ICP-AES). Soil pH was measured from 1:1 soil-water suspensions using a dual probe robotic pH meter equipped with glass electrodes (Skalar Analytical, Netherlands). Buffer pH was determined using the Sikora buffer method. Soil organic matter (SOM) content was determined using the loss on ignition method at 752°F (400°C) and a 4-hour ignition phase.

Results
Correlations between the M3 and DTPA were positive and relatively strong for Zn, Fe, and Cu but weak for Mn ( Figure 1). Simple linear regression models between M3 and DTPA indicated that, on average, micronutrient concentrations were higher in M3 extracts than DTPA extracts. However, more complex regression models suggest these relationships were also influenced by soil pH (Table 2; Figure 2). Using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) values to compare the models provides further weight toward including soil pH, as AIC and BIC values were minimized when soil pH was included in the model for all four micronutrients (Table 2). This pH dependence complicates the interpretation of micronutrients from the Mehlich-3 soil test, as the soil test indices and crop response calibrations currently available were based on the traditional DTPA soil test.

Conclusion
Relationships between M3 and DTPA extractable micronutrient metals were investigated in soils collected from across Kansas. Mehlich-3 and DTPA micronutrients were positively correlated between the two methods. However, Mehlich-3 extracted substantially higher amounts of Cu, Fe, Mn, and Zn in some soils with a high pH. Regression analysis suggests M3 and DTPA soil tests results do not follow "one-to-one" relationships and appear to be influenced by soil pH. Results from this study show that conversion of M3 extractable to a "DTPA equivalent" is possible, but soil pH should be taken into consideration to improve accuracy, especially for Fe and Mn.  Soils with SOM contents exceeding 10% were removed from the study (n = 2). The reported concentrations of copper, iron, manganese and zinc were determined from DTPA extracts.   The effects of soil pH were also taken into consideration. Model fit estimates are illustrated with the shaded ribbons. The soil pH is indicated by color for both soils (points) and models (ribbons); where lighter shades correspond to higher soil pH and darker shades to lower soil pH. Soils with extremely high soil test values were omitted prior to fitting regression models to place focus on the range of agronomic concern (e.g., DTPA-Zn > 10 ppm).