Gitte H. Rubæk

Background Improved use of legacy phosphorus (P) in agricultural soils is requested to reduce the need for P fertilizers. Adapted use of cover crops (CCs) may be a promising tool to support this. Aim We estimated the P allocation to roots and shoots of oat ( Avena sativa , cv Posedion), corncockle ( Agrostemma githago ), and lupine ( Lupinus angustifolius , cv Iris) and their effect on soil enzyme activity, microbial community structure, and indices of plant‐available soil P. Methods We grew the CCs in pots on soils with low‐ and medium‐P status. After 40 days, we measured P, N, and C uptake in shoots and roots; soil microbial C, N, and P; and pH and inorganic P extracted with water (P H2O ) and anion‐exchange resins (P resin ). Soil microbial activity and community structure were assessed by determining phosphomono‐ and phosphodiesterase, β‐glucosidase, and N‐acetyl‐glucosaminidase activity and by extraction of phospholipid and neutral lipid fatty acids (PLFAs and NFLAs). Results Corncockle and lupine took up similar amounts of P, but corncockle had an almost fourfold higher concentration of P. In the low‐P soil, the activity of phosphomonoesterase and soil microbial biomass (total microbial PLFA) were higher after lupine. CCs did not affect P H2O , but after corncockle, P resin was reduced in the medium‐P soil. Oat enhanced the presence of arbuscular mycorrhizal fungi in soil. Conclusions Our results thus suggest that CC species with different P uptake and P uptake strategies can modify aspects in soil of potential importance for the P supply of the following main crop.