Elsevier

Ecological Engineering

Volume 102, May 2017, Pages 166-177
Ecological Engineering

Retention and survival of E. coli in stormwater biofilters: Role of vegetation, rhizosphere microorganisms and antimicrobial filter media

https://doi.org/10.1016/j.ecoleng.2017.02.009Get rights and content

Highlights

  • Retained E. coli concentration decreased with increasing filter media depth.

  • Retained E. coli concentrations in rhizosphere and bulk soils were similar.

  • E. coli die-off rate increased in the presence of rhizosphere microbes.

  • Root exudates from biofilter plants has a net negative effect on E. coli survival.

Abstract

The public health risks associated with pathogens in urban stormwater have been well established, making it necessary to ensure adequate treatment of the stormwater before it is discharged into recreational water bodies or is harvested for reuse. Biofilters, also known as stormwater bioretention systems or raingardens, have shown promising, yet variable, results in reducing indicator bacteria in stormwater. Different biofilter design elements, such as filter media composition and vegetation type, have been found to cause this variable removal performance. Although plants play a key role in the treatment of pollutants, relatively little work has been conducted to understand the importance of interactions between vegetation and the biofilter microbial community on fecal microbial removal. A laboratory-scale biofilter experiment was conducted using Escherichia coli as the test fecal microorganism. Biofilter columns with differing soil media and vegetation types were dosed over a two month span, during which inflow and outflow samples were collected to evaluate system performance. The columns were then decommissioned to collect rhizosphere and bulk soil samples. Root exudates were extracted and used in an E. coli survival study to evaluate their contribution to system performance. The study demonstrated that the antagonistic effects of root exudates/rhizosphere microbes and Cu2+ exchanged zeolite antimicrobial filter media adversely impact the survival of E. coli retained within stormwater biofilters. Furthermore, leaf and flower/seed extracts of L. continentale showed some potential antibacterial activity against E. coli. This work supports the concept that natural processes in biological systems can deliver effective results in the removal of fecal microorganisms, and should be promoted to the extent possible in stormwater green infrastructure.

Introduction

Stormwater has been identified as an emerging alternative water resource, and constitutes an important component of the urban water cycle. However, a wide range of pathogens, present at varying concentrations, pose a significant human health risk when stormwater is harvested for reuse or when individuals are exposed during recreational activities (Haile et al., 1999, Geldreich, 1996, Arnone and Walling, 2007). Hence, the adequate treatment of stormwater, before contact is made with humans during such activities, is essential (NHMRC, 2009).

Biofilters, also known as bioretention systems or raingardens, are soil-plant based systems that promote infiltration and evapotranspiration of stormwater (FAWB, 2009). This technology has shown promising yet variable results in reducing indicator bacteria in stormwater (Hathaway et al., 2011, Zhang et al., 2012). Fecal microorganisms can be sequestered in biofilter media as a result of straining and adsorption during wet weather events (Stevik et al., 2004, Zhang et al., 2010). Subsequently, these captured microbes experience die-off due to the hostile environment prevalent in biofilters, which is characterized by the presence of competitors, predators, solar irradiation, as well as highly variable temperatures (Chandrasena et al., 2014a, Zhang et al., 2012). Biofilter design components, such as filter media type and vegetation, along with operational conditions, such as intermittent drying and wetting, have been found to affect these removal pathways, resulting in the observed variability in fecal microbial removal performance in stormwater biofilters (Li et al., 2012, Chandrasena et al., 2014b). Yet, despite these adverse conditions for survival, a proportion of these microbes can persist during dry weather periods and then be released through desorption during subsequent wet weather periods (Chandrasena et al., 2013).

Media and plant types have been identified to play a major role in nutrient removal in stormwater biofilters (Read et al., 2010, Payne et al., 2014a). While considerable efforts have been made to optimize fecal microbe removal in biofilters, by enhancing adsorption and inactivation using modified filter media (Li et al., 2014, Zhang et al., 2010), relatively little work has been conducted to understand the importance of the interactions between vegetation, the biofilter microbial community, and soil media in fecal microbial removal in stormwater biofilters. Previous studies have reported that several plant species that are present in constructed wetlands used for wastewater/stormwater treatment may have bactericidal properties (Soto et al., 1999, Stottmeister et al., 2003, Vymazal, 2005, García et al., 2010, Malaviya and Singh, 2012). However, most of these studies are based on examination of microbial removal performance in the presence/absence of a given plant species, without further investigation into the underlying mechanisms (García et al., 2010, Chandrasena et al., 2014b). Therefore, the extent to which these plant-related antimicrobial compounds directly affect fecal microbial removal remains poorly understood.

Plant roots are well known to govern the microbial dynamics in terrestrial systems through the root exudation process (Walker et al., 2003, Pinton et al., 2007). These root exudates are comprised of oxygen, sugars, amino acids, and organic acids (Stottmeister et al., 2003). Hence, the rhizosphere has been discovered to harbor a significantly higher number of microorganisms than are found in bare soil (Mukerji et al., 2006, Pinton et al., 2007). However, root exudates may also comprise antimicrobial compounds such as coumaric acid, ferulic acid, and 3-indolepropanoic acid to protect plants from microbial pathogens (Strehmel et al., 2014). Upon being transported into the biofilter media via stormwater runoff, some fecal microbes can attach to the rhizosphere. These fecal microbes may then be exposed to antimicrobial root exudates, adversely affecting their survival. Furthermore, these organisms must compete with other rhizosphere microbes and are exposed to predators. Thus, such root exudates and rhizosphere microbes are hypothesized to have a significant impact on the survival of retained fecal microbes in stormwater biofilters.

Root exudate composition depends on environmental factors such as soil chemistry and the composition of the microbial population (Pinton et al., 2007). As efforts to enhance microbial sequestration and inactivation progressively shift toward the use of modified biofilter media, obtaining a proper understanding of these biochemical interactions is, to an increasing extent, critical to the holistic evaluation of system function. Introduction of novel antimicrobial filter media, such as Cu2+ exchanged zeolite (Li et al., 2014), alters the copper concentration in the biofilters. Soil microbial communities and root exudates in these antimicrobial layers are likely very different to those of a traditional biofilter media, potentially influencing the survival of retained fecal microorganisms to a greater extent than traditional systems. However, to date, no research has been conducted to investigate the effects of “next generation” biofilters (those with antimicrobial filter media) on the interactions between root exudates and microbial communities.

Apart from root exudates, various plant extracts from leaves, flowers, and seeds have also demonstrated antimicrobial activity against fecal microorganisms. Several plant species belonging to the genus Leptospermum are commonly used in stormwater biofilters and, additionally, are recognized for the antibiotic properties of essential oils produced from their leaves/flowers (Demuner et al., 2011) and honeys (Blair et al., 2009). Leaves, flowers, and seeds of biofilter vegetation fall onto the biofilter surface, and eventually decompose into the top media layers during biofilter operation, potentially releasing associated antimicrobial compounds into the biofilter media in the process. As the topmost layers in stormwater biofilters, are where the highest concentrations of retained fecal indicator bacteria are located (Chandrasena et al., 2014a), captured microbes may be exposed to these plant-related antimicrobial compounds. As no research has been conducted to test the antimicrobial activity of biofilter plant extracts, the effect of these compounds on microbe vitality is largely unknown.

In conclusion, little is known as to how the combined effects of biofilter vegetation, rhizosphere microbes, and soil media composition influence indicator bacteria within stormwater biofiltration systems. Specifically, after indicator bacteria are sequestered in biofilters, what processes influence their survival in the filter media? The objectives of this study are: (1) to investigate the distribution of sequestered E. coli in rhizosphere and bulk soils of laboratory-scale biofilters, and (2) to investigate the effect of root exudates, rhizosphere microbes, and various plant extracts on the survival of E. coli in biofilters.

Section snippets

Methods

A laboratory-scale experiment, comprising established (2-year-old) biofilter columns, was conducted in a constructed greenhouse with a clear, impermeable roof that admits full, natural sunlight. E. coli was used to represent fecal microbe sequestration and survival, as it is the one of the commonly used indicator organisms in Australian water harvesting guidelines (NHMRC, 2006, NHMRC, 2009), and is used internationally to evaluate contamination in surface waters. The biofilter columns used in

E. coli removal performance

E. coli removal performance, observed during the single sampling round of the current study, varied among different configurations. The lowest removal performance was observed in WS and PB (average log reduction ∼1), in the current study, while the highest removal was observed in LC and CA (average log reduction >2) (2014 in Fig. 2a). The novel design with antimicrobial filter media achieved only 1.2 log reduction, which was lower than the expected average 2 log reduction shown in the work of

Conclusions

A laboratory-scale study was conducted to investigate the role of biofilter vegetation and rhizosphere microbes in E. coli removal across two types of filter media. E. coli removal performance was found to be dependent on both the type of vegetation and the type of filter media. Stormwater biofilters with a traditional soil composition, planted with either L. continentale or C. appressa, achieved the highest log reduction (>2 log reduction), followed by the next generation layered Cu2+

Acknowledgments

The authors would like to gratefully acknowledge the assistance of Anthony Brosinsky and Richard Williamson throughout the duration of the experiment. This work was financially supported by the Australian Research Council – Discovery Early Career Researcher Award (Grant number DE140100524) and the National Science Foundation – United States (Grant number 1361572).

References (50)

  • R.R.E. Artz et al.

    Survival of

    (2002)
  • S.S. Bhojwani et al.

    Chapter 2 laboratory requirements and general techniques

    Studies in Plant Science

    (1996)
  • S.E. Blair et al.

    The unusual antibacterial activity of medical-grade Leptospermum honey: antibacterial spectrum, resistance and transcriptome analysis

    Eur. J. Clin. Microbiol. Infect. Dis.

    (2009)
  • S.A. Bradford et al.

    Physical factors affecting the transport and fate of colloids in saturated porous media

    Water Resour. Res.

    (2002)
  • S.A. Bradford et al.

    Transport and straining of E. coli O157:H7 in saturated porous media

    Water Resour. Res.

    (2006)
  • T.M. Bright et al.

    Impact of storm-water runoff on clogging and fecal bacteria reduction in sand columns

    J. Environ. Eng.

    (2010)
  • G.I. Chandrasena et al.

    Evaluating Escherichia coli removal performance in stormwater biofilters: a preliminary modelling approach

    Water Sci. Technol.

    (2013)
  • G.I. Chandrasena et al.

    Survival of Escherichia coli in stormwater biofilters

    Environ. Sci. Pollut. Res.

    (2014)
  • Chandrasena, G.I., Deletic, A., Lintern, A., Henry, R., Mccarthy, D.T., submitted. Escherichia coli removal in...
  • K.K.G.I. Chandrasena

    Fecal Microorganism Removal in Stormwater Biofilters

    (2014)
  • S.R. Crane et al.

    Modeling enteric bacterial die-off: a review

    Water Air Soil Pollut.

    (1986)
  • A.J. Demuner et al.

    Seasonal variation in the chemical composition and antimicrobial activity of volatile oils of three species of leptospermum (Myrtaceae) grown in Brazil

    Molecules

    (2011)
  • FAWB

    Adoption Guidelines for Stormwater Biofiltration Systems

    (2009)
  • J.V. Gagliardi et al.

    Intact soil-core microcosms compared with multi-site field releases for pre-release testing of microbes in diverse soils and climates

    Can. J. Microbiol.

    (2001)
  • J. García et al.

    Contaminant removal processes in subsurface-flow constructed wetlands: a review

    Crit. Rev. Environ. Sci. Technol.

    (2010)
  • Cited by (46)

    • Plants against pathogens: Effect of significant antimicrobial-producing plants on faecal microbe inactivation throughout the soil profile of stormwater biofilters

      2022, Water Research
      Citation Excerpt :

      To date, no studies have quantified these effects (Muerdter et al., 2018). Of potentially greater significance is that certain biofilter-suitable plants possess antimicrobial properties with activity against common stormwater faecal microorganisms (Chandrasena et al., 2014b; Chandrasena et al., 2017; Shirdashtzadeh et al., 2017). Indeed, preliminary research by Chandrasena et al. (2014b) indicated that biofilters configured with two Australian plants possessing antimicrobial properties, Melaleuca incana and Leptospermum continentale, achieved 10-fold higher Escherichia coli (E. coli) reductions in treated effluent than other planted systems (irrespective of retention times).

    • Highly variable removal of pathogens, antibiotic resistance genes, conventional fecal indicators and human-associated fecal source markers in a pilot-scale stormwater biofilter operated under realistic stormflow conditions

      2022, Water Research
      Citation Excerpt :

      Characterization of microbial removal in field-scale biofilter systems is challenging due to the highly variable conditions under which these systems operate; specifically, the very large transients in inflow, outflow and media saturation associated with storm and inter-storm periods (Parker et al., 2021). While some microbial removal studies have been performed at the field-scale (Davies et al., 2008; Hathaway et al., 2011; Petterson et al., 2016; Youngblood, 2017), most have been conducted in laboratory-scale microcosms or mesocosms (Afrooz and Boehm, 2017; Mohanty et al., 2014; Mohanty et al., 2013; Mohanty and Boehm, 2014; Mohanty and Boehm, 2015; Chandrasena et al., 2014, 2017b) and in many cases the focus has been on characterizing the removal of fecal indicator bacteria (FIB) such E. coli, enterococci bacteria and fecal coliform or fecal indicator viruses (FIV) such as coliphage, rather than the removal of human pathogens that drive public health risk. Studies of pathogen removal have either utilized laboratory column experiments (Chandrasena et al., 2017a; Graham et al., 2021; Jung et al., 2019; Li et al., 2012) or evaluated the removal of only one pathogen (Campylobacter) at the field scale (Chandrasena et al., 2016).

    View all citing articles on Scopus
    View full text