Estimating Labile and Dissolved inorganic Phosphate Concentrations in Surface Runoff

topographic, and climatic conditions, mathematical Mathematical models are needed to better estimate effects of alternative management methods on the amounts of bioatailable P lost in surface runoff over a variety of soil, topographic, and climatic conditions. I n this study, methods for estimating concentrations of soil-derived labile (dissolbed plus adsorbed) and dissolved P in runoff are proposed and tested. The methods are based on the assumption that most labile P contributed by soil to runoff is derived from soil particles that are detached and transported with runoff water. Effects of erosive selectivity are accounted for by dividing soil and sediment into five undispersed size fractions. Isotherms relating adsorbed labile and dissolved P levels were developed for each from soil analyses. Labile P concentrations in runoff were estimated by summing the product of labile P concentrations prior to entrainment in runoff and measured sediment concentrations for each of the size fractions. Dissolved P concentrations were estimated by distributing labile P between dissolved and adsorbed forms using a mass balance approach. Methods of estimation were tested with runoff from 14 natural rainstorms occurring on 10 fallow plots having soil P levels ranging from about 1.3 to 3.2 mmol kg-' as labile P. Linear regressions of observed on estimated values of both labile and dissolved P were significant. For most events, slopes and intercepts of regression lines were not significantly different from one and zero, respectively, indicating good absolute agreement. Results indicate that soil analyses in combination with models for predicting the amount and size distribution of eroded soil form a basis for estimating movement of bioavailable P forms in runoff. However, additional studies are needed to determine the range of soil types for which the assumptions inherent in the development of the approach are valid. Additional Index Words: bioavailable P, eutrophication, nonpoint-source pollution, aggregates, adsorption isotherms. Wendt, R. C., and E. E. Alberts. 1984. Estimating labile and dissolved inorganic phosphate concentrations in surface runoff. J. Environ. Qual. 13:613-618. The bioavailability of P in runoff is an important consideration in developing management strategies for reducing rates of eutrophication of receiving waters (Sonzogni et al., 1982). Phosphorus forms and their bioavailability have been addressed in recent reviews by Sonzogni et al. (1982) and Nelson and Logan (1982). The P form that is directly bioavailable is dissolved inorganic phosphate (predominantly HPOd2and H,PO,at normal pH's). Other P forms become available through conversion to this phosphate form. In most cases, these forms in runoff either convert to dissolved inorganic P relatively slowly or are usually present in relatively small amounts. However, dissolved inorganic P equilibrates rapidly with P adsorbed on particulate surfaces; hence, the particulates serve as a readily accessible reservoir of bioavailable P. Collectively, the dissolved and adsorbed P are often referred to as labile P. To aid in assessing effects of alternative management strategies on P transport in runoff over a variety of soil, models for predicting P movement have been proposed. Conceptually, these models have addressed interactions of rainfall and runoff water with soil in developing predictive equations, most of which are empirically derived. In some instances, only dissolved P in runoff has been considered (Romkens & Nelson, 1974; Sharpley et al., 1978, 1981), which may neglect significant amounts of bioavailable P on sediments. Models that have addressed both dissolved and sediment-associated P have often attempted to describe only total P quantities on sediments (Davis & Donigian, 1979; Frere et al., 1980), of which only a fraction may become bioavailable in aquatic systems. Other modeling approaches (in most cases developed to predict pesticide movement in runoff) have considered quantities of labile chemical (Frere et al., 1975; Haith, 1980; Steenhuis & Walter, 1980; Leonard & Wauchope, 1980). These approaches have used adsorption isotherms and mass balance considerations to predict amounts of dissolved and adsorbed chemicals in or at the interface of a surface soil zone assumed to interact with rainfall and runoff water. Instantaneous equilibrium between dissolved and adsorbed forms has usually been assumed. Movement of dissolved chemical to runoff has been variously estimated by: (i) assuming concentrations in runoff to equal those in water in the surface soil zone (Frere et al., 1975; Steenhuis & Walter, 1980), (ii) by assuming dissolved chemical is removed from the surface soil zone in proportion to the fraction of rainfall as runoff (Haith, 1980) and, (iii) by assuming the concentration in runoff is equal to that predicted at the interface of the surface soil zone and runoff stream using an empirically derived, effective solution/soil ratio (Leonard & Wauchope, 1980). Adsorbed chemical loss in runoff has been estimated from the amount of soil eroded and the concentration of adsorbed chemical on the eroded soil. The latter has been derived from adsorption isotherms using predicted concentrations of dissolved chemical in the surface soil zone, both with corrections for selective removal of smaller-sized particles (Leonard & Wauchope, 1980; Frere et al., 1975) and without such corrections (Haith, 1980; Steenhuis & Walter, 1980). Approaches similar to the above may be useful for predicting concentrations of dissolved and labile P in surface runoff. However, because P is strongly adsorbed on soil, simplifying assumptions regarding the mechanism of P transfer to runoff that would result in relatively simple predictive equations may be appropriate. This study proposes and tests such approaches for estimating concentrations of soil-derived labile and dissolved P in surface runoff.