Correlation of a resin membrane soil phosphorus test with corn yield and routine soil tests

Most tests for plant-available soil P are based on extraction with a chemical solution. Extraction of P through ion exchange could provide better estimates of plant-available P. The objective of this study was to field calibrate a test based on a commercially available ion-exchange resin membrane (RMP) and compare it with Bray-P, (BP), Mehlich-3 (M3P), and Olsen (OP) tests. Replicated P response trials with corn (Zea mays L.) were established at 59 Iowa locations (78 site-years). The soils represented 17 series of the suborders Argiudolls, Endoaqualfs, Endoaquolls, Hapludalfs, and Hapludolls. Initial BP, M3P, RMP, and OP values ranged from 2 to 63, 7 to 79, 6 to 63, and 3 to 31 mg P kg(-1), respectively. The r(2) values of relationships between soil P extracted across sites ranged from 0.84 to 0.91, and the lowest values were for BP relationships. The BP test measured less P in a CaCO3-affected soil with pH 8.1 but not when pH was less than or equal to7.7. Excluding the calcareous site improved correlations only for BP. Critical concentration ranges defined by Cate-Nelson (CN) and linear-plateau (LP) models and relative yield responses for various soil-test ranges indicated that 13 to 20 mg kg(-1) for BP, 16 to 21 mg kg(-1) for M3P, 13 to 19 mg kg(-1) for RMP, and 8 to 11 mg kg(-1) for OP corresponded to a 94 to 96% mean relative yield response. The sink-based RMP test was as effective as routine P tests at predicting corn response to P fertilization and could be adopted for production agriculture.