Lower and slower: Quantifying phosphorus leaching from struvite relative to monoammonium phosphate by 33P radioisotopic labeling

Abstract

Struvite (NH₄MgPO₄·6H₂O) is a slow-release phosphorus (P) fertilizer hypothesized to reduce nutrient leaching losses relative to highly water-soluble counterparts such as monoammonium phosphate (MAP). We assessed the leaching fate of P from struvite relative to MAP using soil column experiments for two soils representative of the United States Maize Belt, an Arguidoll and a Fragiudalf. Inorganic P and N in leachate (49.8 pore volumes (v_p)) were directly quantified for synthesized ³³P-labeled struvite and ³³P-labeled MAP directly and indirectly by omission method (i.e., by difference with a P-unfertilized treatment). By 8.3 v_p, less than 0.02 % of struvite-P was leached in the Fragiudalf according to both methods of determination, compared to 11.1 % MAP-P determined directly by ³³P tracing and 14.1 % of MAP-P by the omission method. Fertilizer-P leached quantified directly by ³³P labeling was on average 1.2- to 3.7-fold lower than by indirectly by omission for both fertilizers at 2.8–13.9 v_p across both soils, indicating consistent overestimation in P leaching by the commonly used omission method. Leaching of struvite-N was an order of magnitude lower than MAP-N in the Fragiudalf after 8.3 v_p. Struvite entailed 17.4 %–20.4 % lower P leaching and 22.1–46 % lower N leaching compared to MAP across soils under simulated extreme precipitation events, supporting the hypothesized ability of struvite to reduce P as well as N leaching relative to more water-soluble ammonium phosphate fertilizers such as MAP. Additionally, our results point to a methodological need to validate fertilizer P leaching estimated by the commonly used omission method using direct quantification by radioisotopic labeling.