Global phosphorus shortage will be aggravated by soil erosion

Phosphorus (P), crucial for all life, is finite. Unlike nitrogen, which can be endlessly produced through the Haber-Bosch process, P fertilizers originate from non-renewable geological deposits. The one-way flow of P from reserves to farms, freshwater, and oceans has exceeded sustainable limits. While concerns about peak phosphorus have been somewhat alleviated by the discovery of previously overlooked deposits, the socio-economic and political implications of P scarcity remain significant, particularly concerning phosphate imports. The rising global demand has driven up rock phosphate prices. The most significant P loss from ecosystems is via soil erosion, particularly in degraded soils. Most global P cycle evaluations have not accounted for P loss due to soil erosion, despite its acknowledged contribution. This study addresses this gap using spatially explicit data analysis.

Previous studies have highlighted the importance of P loss due to soil erosion, but estimations have been associated with high uncertainty, often relying on extrapolation from regional data or average statistics. MacDonald et al. (2011) calculated that 29% of global cropland had P deficits in 2008. Li et al. estimated a loss of approximately 12.8 Tg yr⁻¹ of P from the world’s croplands. Various studies have attempted to quantify global P loss through different methodologies, yielding a range of estimates. However, most studies haven't included spatially discrete data for analyzing P loss via erosion.

This study combined spatially distributed global soil erosion estimates (considering only sheet and rill erosion by water) with spatially distributed global P content for cropland soils to quantify global soil P loss. The researchers used a 0.5° × 0.5° resolution for the land-use harmonization of land data. The annual organic input data were derived from Ringel et al. (2017), with limitations acknowledged regarding assumptions on maximum input and uncertain atmospheric fluxes. Data for chemical fertilizer, manure, and residue came from Ringel et al. (2017), who based their calculations on data from Bowman et al. (2009), Bormann et al. (2012), and the International Fertilizer Association. Organic P management was defined as the sum of manure and residue input minus plant uptake. Error estimation for calculated fluxes and balances was done using relative, non-symmetric errors to the mean, with error propagation using the standard deviation of total P pool data from Ringel et al. (2017). The RUSLE model was used, with limitations noted regarding its exclusion of gully erosion and landslides. The study indirectly verified predicted P loss by comparing it to river P loads, considering agricultural erosion and runoff contributions, and sediment delivery rates.

The study found that all continents except Asia, Oceania, and Australia showed negative P balances (net P losses from agricultural systems). Even with high chemical fertilizer inputs, P depletion occurred. Africa, South America, and Eastern Europe exhibited the highest P depletion rates. Under a hypothetical scenario of no chemical fertilizer replenishment, global P depletion was projected between 4 and 20 kg P ha⁻¹ yr⁻¹, with water erosion accounting for approximately 60% of total losses. The average P loss due to water erosion from arable soils globally was estimated to be 5.9–17.9 kg ha⁻¹ yr⁻¹. Continental and national regional P losses were between 40 and 85% of total P from agricultural systems, except for Europe and Australia. Regions with intense agriculture or extreme climatic events showed the most dramatic P losses (>20 kg ha⁻¹ yr⁻¹ in areas of eastern China, Indonesia, Africa, Central America, and South America). A comparison with riverine P exports showed that the study's on-site P loss estimates were generally within the range of published riverine P exports, with some regional discrepancies potentially related to differences in methodologies and spatial scales. The study also highlighted high P losses in regions such as eastern China, Indonesia, parts of eastern and southeastern Africa, Central America, and parts of South America.

The findings underscore the vulnerability of the current global land management system, which is heavily dependent on non-renewable P fertilizers. The significant contribution of soil erosion to P loss highlights the need to mitigate erosion to minimize P depletion. The study's comparison with riverine P exports offers a degree of validation, although limitations related to data availability and methodological differences are acknowledged. The results reinforce the urgent need for sustainable soil management practices to reduce P losses and ensure long-term food security.

This study demonstrates the substantial contribution of soil erosion to global phosphorus loss from agricultural soils, even with high fertilizer inputs. Future scenarios with limited fertilizer availability will further exacerbate this problem. Mitigating soil erosion is crucial for sustainable P management, alongside integrated soil fertility management strategies. Further research should focus on refining the spatial and temporal resolution of erosion models and improving the quantification of phosphorus fluxes across various ecosystems.

The study's erosion model considered only sheet and rill erosion, excluding gully erosion and landslides, potentially underestimating total P loss. The indirect verification of P loss using riverine data is subject to uncertainties related to sediment delivery rates and the partitioning of agricultural contributions to river loads. Data limitations and regional variability may affect the generalizability of some findings.