Rain Garden Filter Bed Substrates Affect Stormwater Nutrient Remediation

Abstract

Twelve rain gardens were constructed to analyze the effectiveness of three different filter bed substrates to support plant growth and remove nutrients from urban stormwater runoff. The filter bed substrates included a sand-based substrate (sand) composed of (v/v/v) of 80% washed sand, 15% clay and silt fines, and 5% pine bark; a soil-based substrate (soil) composed of (v/v) 50% sandy loam soil and 50% pine bark; and a slate-based substrate (slate) composed of (v/v) 80% expanded slate and 20% pine bark. Coarse particles (6.3 to 2.0 mm) in the soil-based substrate created a large-pore network that conducted stormwater more quickly into and through the rain garden than slate or sand as evidenced by the high infiltration and saturated hydraulic conductivity values. Sand had good overall retention of pollutants except nitrogen (N) possibly as a result of the very small percentage (5%) of organic matter and low cation exchange capacity (CEC). Soil had the lowest remediation of phosphorus (P) and highest concentration of P in its effluent and was similar in N removal efficiency to slate. Slate had the best retention of N and P. Overall, all three substrates functioned in reducing the quantity of pollutants in urban stormwater runoff; yet, the impact of substrate on remediation appeared to lessen by Season 2 because there were few differences between substrate in the effluent nutrient concentration. Substrate did not affect shoot or root growth. Eleven of the 16 species (B. nigra, B. ‘Duraheat’, M. virginiana, M. ‘Sweet Thing’, I. virginica, I. ‘Henry’s Garnet’, J. effusus, P. ‘Shenandoah’, H. angustifolius, H. ‘First Light’, and E. purpureum subsp. maculatum) grew well in the rain gardens and could be used as rain garden plants.