BIOAVAILABLE PHOSPHORUS DYNAMICS IN AGRICULTURAL SOILS AND EFFECTS ONWATER-QUALITY

The transport of bioavailable phosphorus (BAP, i.e., algal available) in agricultural runoff can accelerate freshwater eutrophication. As pr ocedural and theoretical limitations have restricted BAP estimation in agricultural soils and runoff, a routine method was developed using i ron oxide-impregnated paper strips (Fe-oxide strips) as a sink for BAP . Fe-oxide strips were used to investigate the amounts and seasonal dy namics of BAP in 12 Oklahoma soils over 2 years and the effect of agri cultural management on BAP loss in runoff from these 12 and 8 addition al sites over 4 years. The sites involved native grass, wheat (Triticu m aestivum L.), sorghum [Sorghum bicolor (L.) Moench.], and peanut (Ar achis hypogaea L.). The strip P content of unfertilized soils was high er in winter (October-March; 10-19 mg kg-1) than spring months (May-Ju ne; 3-6 mg kg-1). For the P-fertilized soils, the strip P content incr eased from 8-17 mg kg-1 prior to P application (20-25 kg ha-1 yr-1) to 33-42 mg kg-1 immediately following application. Seasonal dynamics of strip and organic (OP) and inorganic P (IP) fractions indicated miner alization of moderately labile OP, as a function of phosphatase enzyme activity, and release of moderately labile IP and fertilizer P, as a function of P sorption capacity; these are the major processes control ling the strip P content of unfertilized and fertilized soil, respecti vely. Dissolved organic P did not reduce the efficiency of IP sorption by the Fe-oxide strips and removal of P from the strips by NaOH rathe r than H2SO4 minimized hydrolysis of OP sorbed on the strip. The loss of BAP in runoff was a function of watershed management. Over the 4-yr study, BAP losses increased in the order reduced till (98 g ha-1 yr-1 ), native grass (160 g ha-1 yr-1), no till (382 g ha-1 yr-1), and conv entional till (678 g ha-1 yr-1). Although the total P loss was 93% low er from no till than conventional till, 73% was bioavailable for no ti ll compared to only 28% for conventional till. Clearly, strip P is a d ynamic soil property, which, along with watershed management, can infl uence the bioavailability of P loss in runoff. Thus, the BAP content o f runoff in conjunction with runoff and erosion potential may provide a more reliable index of the trophic response of receiving water bodie s than dissolved or total P.