Phosphorus mining for ecological restoration on former agricultural land

Abstract

To restore species‐rich terrestrial ecosystems on ex‐agricultural land, establishing nutrient limitation for dominant plant growth is essential because in nutrient‐rich soils, fast‐growing species often exclude target species. However, N‐limitation is easier to achieve than P‐limitation (because of a difference in biogeochemical behavior), biodiversity is generally highest under P‐limitation. Commonly used restoration methods to achieve low soil P‐concentrations are either very expensive or take a very long time. A promising restoration technique is P‐mining, an adjusted agricultural technique that aims at depleting soil‐P. High biomass production and hence high P‐removal with biomass are obtained by fertilizing with nutrients other than P. A pot experiment was set up to study P‐mining with Lolium perenne L. on sandy soils with varying P‐concentrations: from an intensively used agricultural soil to a soil near the soil P‐target for species‐rich Nardus grassland. All pots received N‐ and K‐fertilization. The effects of biostimulants on P‐uptake were also assessed by the addition of arbuscular mycorrhiza ( Glomus spp.), humic substances or phosphate‐solubilizing bacteria ( Bacillus sp. and Pseudomonas spp.). In our P‐rich soil (111 µg P Olsen /g), P‐removal rate was high but bioavailable soil‐P did not decrease. At lower soil P‐concentrations (64 and 36 µg P Olsen /g), bioavailable soil‐P had decreased but the P‐removal rate had by then dropped 60% despite N‐ and K‐fertilization and despite that the target (<10 µg P Olsen /g) was still far away. None of the biostimulants altered this trajectory. Therefore, restoration will still take decades when starting with ex‐agricultural soils unless P‐fertilization history was much lower than average.