Phosphorus in agricultural soils: drivers of its distribution at the global scale

Abstract

Phosphorus (P) availability in soils limits crop yields in many regions of the World, while excess of soil P triggers aquatic eutrophication in other regions. Numerous processes drive the global spatial distribution of P in agricultural soils, but their relative roles remain unclear. Here, we combined several global data sets describing these drivers with a soil P dynamics model to simulate the distribution of P in agricultural soils and to assess the contributions of the different drivers at the global scale. We analysed both the labile inorganic P (P ), a proxy of the pool involved in plant nutrition and the total soil P (P ). We found that the soil biogeochemical background corresponding to P inherited from natural soils at the conversion to agriculture (BIOG) and farming practices (FARM) were the main drivers of the spatial variability in cropland soil P content but that their contribution varied between P vs. P . When the spatial variability was computed between grid cells at half-degree resolution, we found that almost all of the P spatial variability could be explained by BIOG, while BIOG and FARM explained 38% and 63% of P spatial variability, respectively. Our work also showed that the driver contribution was sensitive to the spatial scale characterizing the variability (grid cell vs. continent) and to the region of interest (global vs. tropics for instance). In particular, the heterogeneity of farming practices between continents was large enough to make FARM contribute to the variability in P at that scale. We thus demonstrated how the different drivers were combined to explain the global distribution of agricultural soil P. Our study is also a promising approach to investigate the potential effect of P as a limiting factor for agroecosystems at the global scale.

Integrative Crop‐Soil‐Management Modeling to Assess Global Phosphorus Losses from Major Crop Cultivations

“…The P tot /P in ratio estimated in this study falls in the middle of the range of previous studies going from between 2.8% and 23% (Table 2). A similar CV was reported by Ringeval et al (2017) in modeling farm P inputs and outputs with a global Figure 5. A similar CV was reported by Ringeval et al (2017) in modeling farm P inputs and outputs with a global Figure 5.…”

“…This necessitates considering plant-soil-management as a coupled system. However, P losses and fertilizer requirements vary substantially among different crops (Ringeval et al, 2017). Liu et al, 2008;Quinton et al, 2010).…”

Quantifying Nutrient Budgets for Sustainable Nutrient Management

“…A simplified two-pool (labile and stable pools) P model, the Dynamic Phosphorus Pool Simulator (DPPS), has been developed to estimate P legacies from regional to global scales (Lun et al, 2018;Mogollón et al, 2018;Zhang et al, 2017). Other researchers used soil P dynamics models with more than two P pools (Goll et al, 2012;Ringeval et al, 2017;Wang et al, 2009;Yang et al, 2013) to study the distribution of soil P. Due to lack of data and understanding, it is difficult to compare and validate soil P pool estimates from different models.…”

Quantifying the Limitation to World Cereal Production Due To Soil Phosphorus Status

“…Since the crop P demand is calculated from the ORCHIDEE‐CROP biomass pools in irrigated conditions, the climate determines the yield potential and subsequently the distribution shape. For potential root P uptake, we have shown that 80% of the spatial variability is explained by the P ILAB distribution (section 3.1), which in turn depends mostly on the historical farm input/output balance and soil biogeochemical background (Ringeval et al, ).…”

“…Because the potential P uptake is common to all crops/methods, this points to the uptake as the main determinant of the yield gap uncertainty. Since the uncertainty in the potential root P uptake is largely explained by P ILAB (Figure , bottom), this uncertainty can be traced back to the uncertainty of the main drivers in the P ILAB pool, which are the history of farm inputs/outputs and the soil biogeochemical background (Ringeval et al, ). For rice, there is a notable difference between the three limitation methods (Figure , top right) that is due to differences between the methods to estimate the demand (as explained in section 3.2).…”