In a joint project between two agencies in Singapore, the National Parks Board and the National University of Singapore—Singapore Delft Water Alliance, more than 30 plants species were screened and tested to select those that are suitable for application as vegetation in bioretention systems. The research project investigated the remediation capacity of the plants and their associated rhizosphere microbial communities. Of the numerous stormwater pollutants, the phytoremediation study focused on nitrogen.

The experiments were set up and conducted from September 2010 through June 2011 at a site in Singapore, with an average day time temperature of 32 ± 1.87 °C and night time temperature of 25 ± 1.28 °C.

The bioretention system comprised of three distinct layers: filtration layer, transition layer and drainage layer. The top filter media composed of 50 % Singapore’s Approved Soil Mix (ASM) and 50 % medium to coarse sand. The transition media is made up of coarse sand and the bottom drainage media is made up of fine gravel. The filter media had a hydraulic conductivity of ca. 136 mm h−1, which was compliant with the 50–200 mm h−1 range proposed in the bioretention design guidelines (PUB 2009).

Table 1

To test the remediation capabilities of the plants for nutrients, irrigation water was chemically spiked to give a final concentration of 10 mg L−1 nitrate and 2 mg L−1 phosphate. These nutrient concentrations were above the levels commonly detected in urban stormwater runoffs, particularly in Singapore (Chua et al. 2009).

At the end of the experiment, the bioretention setups exhibited 100 % efficiency in removing phosphate. However, phosphate was also completely removed in the unplanted control, indicating that the remediation of phosphate was primarily attributable to the bioretention substrate.

Eleven plant species demonstrated high efficiency in nitrate uptake, removing more than 85 %. The plant information of the 24 species with more than 50 % nitrate uptake level is presented below.

Table 2