Impact of soil composition on nitrate and phosphate removal efficiencies: a bioretention mesocosm study

Abstract

While much is known about the design criteria and application of rain gardens and bioretention systems, their effectiveness for nitrogen and phosphorus removal is variable and inconsistent. Some studies have suggested that this variability is due to differences in soil composition. The goal of this work was to evaluate how heterogeneous (sandy and sandy loam) and homogeneous (ASTM C-33 sand only) soil mixtures impacted nitrate and phosphate removal in vegetated and non-vegetated bioretention mesocosms (RT = 24 h). Vegetated mesocosms for each soil treatment contained three drought and water tolerant native plant species, while non-vegetated mesocosms contained soil only. Overall, N & P removal was more variable in the heterogeneous mesocosms than in homogeneous mesocosms. Net nitrate removal in vegetated heterogeneous mesocosms was at least 80% and in non-vegetated mesocosms concentrations increased by at least 66%, while phosphate removal was negligible in both vegetated and non-vegetated mesocosms. Conversely, for the vegetated homogeneous mesocosms we observed significant and consistent nitrate (87%) and phosphate (100%) removal efficiencies. For the non-vegetated homogeneous mesocosms, removal efficiencies were lower and more variable, 66% nitrate and 25% phosphate removed. These results demonstrate that bioretention mesocosms containing ASTM C-33 sand media perform better than soils containing organic materials, reducing the quantity of nitrate and phosphate in the system. Collectively, these results show that nitrate and phosphate removal in bioretention systems is highly dependent on both soil type and the presence or absence of plant material suggesting that soil composition is critical for maximizing the nutrient removal efficiencies in rain gardens and bioretention systems.