Phosphorus sorption dynamics in soils and coupling with surface and pore water in riverine wetlands

Adsorption to soils is one of the dominant mechanisms of P storage in wetla nds. We examined P sorption dynamics in soils collected at 12 sample points with diverse hydrology, geomorphic position, mineralogy, and plant communi ties in two riverine wetlands in northern Minnesota and Wisconsin. Phosphor us sorption parameters from these 12 sample points were correlated with cor responding biogeochemical variables and subsequently extrapolated across 15 7 sampling points in the two wetlands, based upon a large spatial dataset, We then used a series of single and stepwise regressions to determine the b est set of predictive variables for surface water, soil, and plant P pools. Intrasite variation in P sorption dynamics was greater than intersite vari ation between the two wetlands and rivaled the variation found in the liter ature for both upland and wetland soils. An essentially constant final P co ncentration occurred st moderate P additions (less than or equal to 32 Pmol P L-1), indicating extreme soil buffering capacity of porewater P concentr ations. Spatial variation in soil P pools across each wetland were predicte d very well in stepwise regressions, particularly in the summer (R-2 = 0.49 -1.00). Variables that were important in explaining this variation included the amount of P sorbed at equilibrium, maximum P sorption capacity, percen tage of P sorption sites occupied at equilibrium, organic matter content, b ulk density, and oxalate-extractable Fe and AL content. Phosphorus concentr ations in surface water were predicted less well by stepwise regression (R- 2 = 0.04-0.46), suggesting only weak-to-moderate spatial coupling between s oils and surface-water P dynamics. Plant P pools were predicted poorly. Om results indicate the importance of geochemical sorption in controlling P dy namics in riverine soils. We suggest that nutrient studies in spatially div erse wetlands must be designed in a manner that adequately captures the ric h spatial dynamics of the system.