Abstract
Portable X-ray Fluorescence probe (pXRF) is a tool used to measure many elements quickly and efficiently in soil with minimal sample preparation. Although this sensing technique has been widely used to determine total elemental concentrations, it has not been calibrated for plant-available nutrient predictions. We evaluated the potential of using pXRF for fast plant-available nutrient quantification. Two experiments were conducted in soils treated with two types of biochars to obtain a practical range of soil pH (5.5 - 8.0) and organic carbon (2.0 - 5.5%). Biochars applied were derived from switchgrass (SGB) and poultry litter (PLB). The first experiment received biochars at application rates up to 8% (w/w) and had no plants. The second experiment had up to 4% of SGB or PLB planted with ryegrass (Lolium perenne). Linear regression (LR), polynomial regression (PolR), power regression (PowR), and stepwise multiple linear regression (SMLR) were the models tested. Regardless of the extraction method, phosphorus (P) showed a strong relationship between pXRF and several laboratory extraction methods; however, K prediction via pXRF was sensitive to the plant factor. The optimum soil available-P corresponding to the maximum P uptake in plant tissues can be assessed with pXRF. The LR was inconsistent for calcium (Ca), sulfur (S), and copper (Cu) and non-significant for magnesium (Mg), iron (Fe), and zinc (Zn). Our results showed that pXRF is applicable to estimate P availability in soils receiving organic amendments. More evaluations are needed with diverse soil types to confirm the findings before using pXRF for fertilizer recommendation.