Comparing soil testing methods for soil organic matter, lime requirements, and developing a phosphorus soil test correlation

Abstract

The Kansas State University Soil Testing Laboratory currently uses the Walkley-Black (WB) method for soil organic matter (SOM) estimations, the Shoemaker-Mclean-Pratt (SMP) buffer for lime recommendations, and bases the soybean phosphorus (P) critical value for P fertilizer recommendations off other crops. Hazardous waste is produced from WB and SMP creating a health hazard for workers, and substantial cost for handling and disposal. The substantial increase in land area devoted to soybean creates the need to validate currently assumed soil test P critical value and check the current P recommendations for that crop. Overarching objectives of this dissertation are to find suitable non-hazardous replacements for WB and SMP, and to find the soybean P critical value in Kansas. Three common methods used to estimate SOM are WB, dry combustion (DC), and loss on ignition (LOI). An experiment was set up using 98 Kansas soils to compare WB, scooped and weighed, LOI scooped, and DC weighed. All methods correlated well to each other with LOI to weighed WB, LOI to DC, and WB weighed to DC, having correlation coefficients of 0.97, 0.98, and 0.98, respectively. The lowest variability was observed with DC, followed by WB weighed, LOI, and then WB scooped with average standard deviations of 0.04, 0.13, 0.17, and 0.24, respectively. Two non-hazardous alternatives to the SMP buffer to determine soil lime requirement are the Sikora buffer, and the modified-Mehlich buffer. Sikora’s buffer is designed to mimic SMP. Buffer values alone or Mehlich’s equation may be used to calculate lime requirements. Thirty seven soils with a pH less than 5.8 were incubated at lime rates 0, 2240, 4480, 8960, and 17920 kg ECC ha[superscript]-1. Amount of lime required to reach pHs 6.0, 6.3, and 6.6 was calculated. Mehlich’s equation better predicted lime requirements for all target pHs and buffers than buffer pH alone. The Sikora buffer with Mehlich’s equation provided a better lime estimation than the Mehlich buffer using Mehlich’s equation. A P correlation and calibration study was conducted with soybeans at 23 sites in Eastern Kansas from 2011 to 2014. Soil Mehlich-3 P available P was compared to relative soybean yield at these sites. Soybean P critical value was found to be between 10 and 15 or 11.6 mg kg[superscript]-1 using Cate-Nelson, and linear-plateau models, respectively. A linear response to P and relative yield was observed on soils testing between 3 and 8 mg kg[superscript]-1, but not on higher testing soils.