Factors Affecting Mehlich III Soil Test Methodology for Extractable P

Agronomic and environmental testing laboratories in Texas and elsewhere have adopted Mehlich III (M3) as their official soil test phosphorus (P) methodology. However, M3-P data could be skewed due to non-homogenous soil samples or failure to follow standard protocol which could influence recommendations or restrictions. Twelve agricultural soils with a wide range of properties were collected from across Texas. Exhaustive efforts via multiple methods were made to prepare homogeneous representative soil samples. The standard M3 protocol selected was a 2 g weighed soil sample placed in a 148 ml disposable plastic cup, using a 1:10 soil:M3 solution ratio, shaken on a 200 rpm orbital shaker with a 2.5 cm throw for exactly 5 min, and filtered through Whatman No. 2 filter paper. The standard protocol was compared with nine different protocol variations with variables including soil weighing versus scooping, scooping repeatability of different technicians, soil sample weight, shaking type, speed and time, different filter papers, and varying soil:extractant ratios. Extent of soil pulverization on M3-P results was also evaluated. Tests were performed in four replications for all protocols to assess effects on M3-extractable soil P. Percent recovery of soil during grinding had no effect on M3-extractable P. Little difference in M3-extractable P was observed between scooping and weighing of 2 g soil samples. Shaker type had no effect on extractable P in soils with low clay contents, however, increasing shaking speed and using an orbital shaker resulted in higher extractable P, especially in clayey soils. Both Whatman No. 1 and 2 filter papers were found suitable for M3-P analyses. Different soil:extractant ratios resulted in a highly significant influence on the amount of M3-P extracted. However, when ratios were maintained between 1:9 and 1:11, few differences in extractable P were observed. Using sample weights below 3 g did not significantly alter precision or accuracy of results. However, technician variation in scooping of 2 or 5 g soil samples resulted in significant differences in M3-P.