Comparison of Soil Phosphorus Extractants as Predictors of Mycorrhizal Dependency

Abstract

AbstractTo determine which of several soil P extraction techniques best predicted the response of citrus (Poncirus trifoliate X Citrus sinensis) seedlings to mycorrhizal fungi (Glomus fasciculatus), 25 nursery and field citrus soils from southern and central California were collected and analyzed for available P by five methods. Mycorrhizal dependency (MD), defined as the dry weight (mycorrhizal plants)/dry weight (nonmycorrhizal plants) × 100, was significantly correlated with the reciprocal of extractable soil P regardless of the extraction method. Those methods whose results depend most on soil solution P concentrations gave best results. Saturation extract P(PSE, R2 = 0.67***), anion exchange resin P(PAER, R2 = 0.57***), and 1:10 soil to water extract P(P1:10, R2 = 0.51***) were all acceptable for predicting mycorrhizal response. Bicarbonate‐extractable P (PB1C, R2 = 0.32**) and ammonium fluoride P (PAF, R2 = 0.19*) were less acceptable. Using regression models, based upon the reciprocal of the extractable‐P value, a MD > 200% was predicted where PSE < 0.65 mg L−1, PAER < 62 mg kg−1 soil, P1:10 < 0.66 mg L−1, PBIC < 34 mg kg−1 soil, and PAF < 54 mg kg−1 soil. Where analytical capabilities are adequate, PSE is the preferred method for predicting response, while PAER would suffice where higher analate P concentrations are required. Alternatively, inclusion of both PSE and PAER gave an excellent correlation with response (R2 = 0.82***). Inclusion of DTPA‐extractable Mn significantly improved regression equations in all cases except that of PSE, where DTPA‐extractable Zn significantly improved the relationship.