Soil meets stream: Vertical distribution of soil phosphorus in streambanks

Characterizing the vertical distribution of total phosphorus (P) and P forms in streambanks can improve estimates of P loads to streams via bank erosion. Using a systematic literature review, this meta-analysis evaluated the vertical distribution of total P (P_tot), Bray-1 test P (P_Bray1), oxalate-extractable P (P_ox), degree of P saturation (DPS_ox), and soil P storage capacity (SPSC) at 161 streambank locations encompassing diverse soil orders and riparian land covers. Across studies, maximum soil sampling depth (125 cm) was on average half that of streambank height (250 cm on average), and water quality-related P fractions such as water-extractable P (P_w) were not generally reported. Slightly weathered Entisols and Inceptisols collectively accounted for 84% of observations, likely reflecting alluvial deposition environments along stream corridors, and exhibited decreasing P_tot (820 to 400 mg kg⁻¹) with increasing depth to 200 cm. Irregular changes in P_tot at depth > 200 cm of Entisols and Inceptisols suggested burial of organic-rich A horizons by alluvial deposition. Changes in riparian land cover mainly affect P_tot in the upper 60 cm, with depth-weighted mean P_tot increasing as grassland (217 mg kg⁻¹) < forest (628 mg kg⁻¹) ≈ wetland (624 mg kg⁻¹) < developed (684 mg kg⁻¹) < agriculture (739 mg kg⁻¹), consistent with expected anthropogenic P enrichment of surface soils and posing increased P loading risk. Effect size analysis showed that riparian land cover had a large influence on P_tot within the top 30 cm depth whereas soil order and bedrock type had large influences on subsurface P_tot, suggesting depth-divergent drivers of P in streambank soils. Additionally, P_tot and P_ox were positively correlated with silt, clay, carbon, cation exchange capacity (CEC), oxalate-extractable iron (Fe_ox) and aluminum (Al_ox), and total Fe (Fe_tot) contents. This work helps advance understanding of the vertical variation of streambank soil P across diverse soil types, land covers, and soil properties, and their implications for P loading from bank erosion. We also identify gaps in sampling approaches (insufficient depth) and analysis (P_w) that may be hindering comprehensive assessment of bank erosion P loading and risk to water quality. Finally, we illustrate how fluvial erosion and deposition may generate complex vertical P distributions in streambanks, with implications for estimating erosional P loading that can benefit from a pedologically explicit approach.