Natural treatment systems are divided into several categories: land-treatment systems, wetland systems (both for stormwater treatment and wastewater treatment), aquatic systems (ponds and floating aquatic plant systems), and on-site and small community systems in this review.
LAND TREATMENT SYSTEMSSediment transport was studied in non-submerged overland flow over grass by Deletic (1999). Sediment concentration in runoff was found to decrease exponentially with distance and reaches a constant value. A simplified relationship between particle fall number, Nf, and the percentage of particles trapped in the grass was established. Munoz-Carpena et al.(1999) described a singleevent model to predict sediment transport in vegetative filter strips. The model links three submodels: a Petrov-Galerkin finite element kinematic wave overland flow submodel, a modified Green-Ampt infiltration submodel, and the University of Kentucky sediment filtration model. A set of 27 natural runoff events were used to test the model, with good predictive capability. A study of phosphorus storage on lands irrigated with treated sewage effluent for over 20 years found that P accumulation occured primarily in organic forms, both within the soil and as peat on its surface (Menzies et al., 1999). This finding throws into question the common practice of predicting disposal scheme life on the basis of P adsorption curves.
WETLAND SYSTEMSGeneral. Gerba et al. (1999) evaluated a duckweed-covered pond, a subsurface flow wetland, and a surface flow wetland for the removal of indicator bacteria and pathogenic protozoa. Larger microorganisms (Giardia and Cryptosporidium) were best removed in the the pond (sedimentation), while the best fecal coliform removal occurred in the subsurface flow wetlands (98%). These results suggest that combinations of ponds and wetlands work best to remove pathogenic organisms. A new mathematical model, using transport detention time, was developed to describe BOD removal in wetlands by Chen et al. (1999). The model treats the wetland as a series of continuously stirred tank reactors and an nth-order ordinary differential equation was derived based on mass balancing and convection/dispersion. The model is compared to existing plug-flow models and is shown to be more accurate, particularly in transient conditions. Xue et al. (1999) quantified denitrification in constructed wetlands receiving agricultural tile drainage. Nitrate disappearance rates in the water column ranged from 12 to 63 mg N m -2 hr -1 .Stormwater Wetlands. Hydrologic design considerations for stormwater treatment wetlands were described by Koob et al. (1999). Detention structure designs based on the time between events was encouraged to account for evaporation, infiltration, and treatment. Revitt et discussed the use of constructed wetland systems to reduce the impacts of urban runoff on receiving stream quality. Sedimentation maintenance and an overflow mechanism to prevent sediment remobilization were described as key elements. A study of vegetative syste...