Environmental Soil Phosphorus Threshold: Is it Affected by Soil Dataset?

Abstract

Degree of phosphorus saturation (DPS) and soil phosphorus storage capacity (SPSC) have been tested worldwide to predict phosphorus environmental risk in agricultural systems. Our study was carried out in Oxisols from Southern Brazil, to estimate the environmental P-threshold and compare the long-term manured soils with unmanured soils under native vegetation (without anthropic interference) incubated with mineral fertilizer. The long-term manured experiments were conducted under no-tillage with dairy liquid manure (DLM) applied at two sites (sandy clay loam texture, since 2005; and clayey texture, since 2006). Soil samples, from the same type as the long-term manured soils, but under native vegetation were incubated with mineral fertilizer for 30 days. DPS was obtained by P, Fe, and Al extracted with ammonium oxalate and Mehlich-1. P-water and P-CaCl₂ were determined to represent the soluble P. Long-term application of DLM affected the environmental P-threshold. The dataset from long-term DLM without incubation resulted in an environmental P-threshold (P Mehlich-1) of 178 mg kg⁻¹ (26% DPS) in the sandy clay loam soil and 77 mg kg⁻¹ (7% DPS) in the clayey soil. Using dataset from soils under native vegetation incubated with mineral fertilizer with liming, the environmental P-threshold (P Mehlich-1) was 75 mg kg⁻¹ (10% DPS) in the sandy clay loam soil and 110 mg kg⁻¹ (13% DPS) in the clayey soil. The distinguished results of environmental P-threshold from the same soil type and texture are a consequence of the soil phosphorus dataset. Datasets constituted by most of values with low DPS or low soil test P (STP Mehlich-1) resulted in low environmental P-threshold, and vice-versa. In general, SPSC was less restrictive than DPS or STP in estimating P environmental risk. It is recommendable to define and follow a protocol to assess P-thresholds creating DPS datasets that provide a high and similar amplitude among different soil types, so the data among places can be compared.