| Soil colloids, with their large specific surface area, high absorption capacity, and abundant surface functional groups, play a vital role in biogeochemical cycles (eg. speciation, transport and transformation) of contaminants. Colloid-facilitated phosphorus (P) transport has been a key mechanism of agricultural non-point source pollution of P, with the premise of colloidal P (Pcoll) mobilization at soil-water interfaces. This study targeting soil Pcoll, investigated size distribution of aquatic P in typical sections of East Tiao River, and Pcoll release potential in soils cultivated with rice (RS), tea (TS), vegetables (VS), nursery stock (NSS), and bamboo (BS) located in representative basin (Caoqiaoxi Basin) of East Tiao River, along with the relationship between Pcoll release potential and P fractions in studied soils. Then locally typical VS and RS samples were selected to further characterize P speciation in soil colloids and bulk soils at the molecular level, and differences of P species in soil colloids and bulk soils were discussed. Considering the studied area is significantly impacted by soil acidification, the release potential of soil Pcoll under different pH conditions and involved mechanism were explored. Finally, efficiency of polyacrylamide (PAM) in controlling release potential of Pcoll was researched. This work should enrich current development on speciation, release mechanism and control of soil P. Detailed results are shown below.In typical sections of the East Tiao River, results indicated particulate P (0.020-0.492mg L-1) was dominant followed by the truly dissolved P (0.034-0.109mg L-1), while Pcoii (0-0.025mg L-1) was quantitatively the lowest in the whole river. From upstream to downstream, particulate total P (TP) and molybdate reactive P (MRP) increased sharply, along with the decrease of truly dissolved TP and MRP. However, colloidal TP (TPcoll) and colloidal MRP (MRPcoll) remained at a relatively stable level in the whole river. The partition coefficient of MRP in colloidal phase declined as colloidal particles increased, which indicated the particle concentration effect. These observations, in this large-scale field investigation, fitted the "colloidal pumping" hypothesis. It may be concluded that colloids act as the intermediate and buffer in the dynamically balanced transfer of P from truly dissolved phase to large particulate phase, having a significant role in size distribution of aquatic P.Release potential of Pcoll was highest in VS and BS, followed by RS and NSS, and was lowest in TS. Colloidal P was the major species of the released P (<1μm), ranging from55.1%to80.9%, for all soil samples under various land use except for TS. Colloidal P mainly consisted of MRP (59.6-96.8%) in all studied soils. P fractions with various labilities in soil samples distributed differently. For all samples, the majority of P was found in the moderately labile fraction (NaOH-P,41-53%). The second largest pool for RS, VS and BS was the labile P fraction (NaHCO3-P and H2O-P), but for TS and NSS it was the nonlabile fraction (HC1-P and residual P). Release potential of MRPcoll presented a highly significant correlation with concentration of H2O-P, NaHCO3-P;, and NaOH-Pt, and a significant correlation with NaOH-Pj and HCl-P, while colloidal MUP did not show correlation with any soil P fractions.Taken RS and VS as examples, molecular speciation of P in soil colloids and bulk soils were further identified by synchrotron-based P K-edge X-ray near edge structure (XANES) and solution P-31nuclear magnetic resonance (P-NMR) spectroscopy. Results showed that P was mainly presented as iron-associated P (Fe-P) in both studied soil colloids (61.1-70.4%) and bulk soils (64.6-65.3%), which explained the significant correlation between release potential of MRPcoll with concentration of soil NaOH-Pj. Increased proportion of hydroxyapatite (HAP) and decreased percentage of soluble calcium-associated P occured in soil colloids as compared to bulk soils. The RS colloids have higher percentage of aluminum-associated P as well as lower ratio of orthophosphate diesters to orthophosphate monoesters than the bulk soil. This indicated stable P species accumulate in soil colloids as compared to bulk soils.The release potential of Pcoll was rather low at pH～3.0for both VS and RS, while high pH (-7.0) significantly enhanced the release of soil colloids and Pcoll, accompanied with the gradual decrease of zeta potential of colloids, which implied the increased interparticle electrostatic repulsion contributed to the release of soil colloids and Pcoll. Colloidal TOC accounted for98.8%and99.3%of the variation of MRPcoll in VS and RS respectively, which allowed for the hypothesis that MRPcoll was probably presented as organic matter (OM)-metal (M)-phosphate complex. The XANES results strongly suggested that MRPcoll occurred as OM-calcium-MRP at pH-5.0, and transformed to Fe-P and HAP at pH～7.0.Application of PAM significantly reduced the release potential of colloids and Pcoll in studied VS and RS.0.1%PAM treatment was recommoned to reduce release of colloids and Pcoll in VS. Contrastly, the controlling effect of PAM on the release potential of colloids and Pcoll in RS increased with higher PAM application rate, and0.2%PAM treatment reduced75.0%of colloid release and88.1%of Pcoll release respectively. The controlling effect of PAM on MRPcon showed similar characteristics with that on Pcoll. No significant inhibition but a possible facilitation effect of PAM on truly dissolved P were shown, which related to land use of studied soils and application rate of PAM. |
PDF Full Text Request