Coliform Bacteria: The Effect of Sediments on Decay Rates and on Required Detention Times in Stormwater BMPs ()
1. Introduction
The presence of fecal indicator bacteria (FIB) in surface waters can have detrimental consequences to human and economic health. For example, exceedingly high occurrences of these bacteria in recreational beaches can put bathers at risk of contracting illnesses. As a prevention to illnesses, beaches are closed, resulting in economic losses in municipalities that depend on tourism [1,2]. To protect recreation and ecosystems in water bodies, the United States’ Clean Water Act 303(d) impaired waters list includes thousands of creeks, rivers, and coastal zones classified as impaired for pathogens [3]. The denotation of “pathogens” includes known pathogenic organisms but, more commonly, FIB, such as total coliforms, fecal coliforms, Escherichia coli (E. coli), and enterococci bacteria. A large contributor to FIB in surface waters is stormwater runoff [4,5]. Solutions to controlling pollution to surface waters from stormwater runoff are called Best Management Practices (BMPs), also known as Stormwater Control Measures (SCMs), and include a range of strategies from educating the public to cleaning animal feces from paved surfaces to constructing stormwater treatment detention basins and wetlands. However, current efforts to reduce concentrations of FIB using stormwater BMPs have yielded inconsistent results. Monitoring efforts of BMP inlet and outlet points during storms generally report minor decreases or even increases of FIB at the outlets [6]. Some monitoring efforts of the same type of BMP but in different locations yield opposite results [7], and some BMPs perform well in all seasons except summer [8].
Reasons for inconsistent results in treating FIB with BMPs are the complexities associated with this type of pollutant. The experiments reported in this paper give insight on what may be necessary for structural BMPs, such as extended detention basins and constructed wetlands, to effectively reduce FIB. An extended detention basin is a constructed pond used to detain excess stormwater runoff as well as promote settling of various pollutants found in runoff. As the stormwater runoff is detained in the basin, sedimentation promotes the reduction of total suspended solids (TSS) and other pollutants attached to suspended solids. A constructed wetland is used in a similar manner to the extended detention basin, but it uses plants to aid in the reduction of TSS, FIB, organic materials, and nutrients in the water.
Bed sediments have the potential to increase FIB concentrations in the water column by furnishing nutrients and additional FIB but may decrease water column FIB by providing a source of predator organisms. In this study we examined how FIB concentrations behave in the water column when bed sediments are present. We hypothesized that the decay rate of FIB in the water column, and therefore in structural BMPs, decreases with the presence of bottom sediments. We used microcosm studies to experimentally determine FIB decay rates for five treatments that provided input into reasons for FIB decay in the water column and the effect of sediments. Further, we estimated the hydraulic residence time necessary in an extended detention basin and constructed wetland to reduce total coliform concentrations to an acceptable level, based on the decay rates from the experiments.
This research builds on literature that considers the impact of sediments and associated organic matter and nutrients on water quality. Organic matter and nutrients in the water column can impact the occurrence of FIB, whether the nutrients come from sediment or another source. For example, Smith and Prairie [9](Smith and Prairie 2004) studied the effect of dissolved organic carbon (DOC) and nutrients on the behavior of bacteria in different types of lakes. They found that bacteria growth rates were positively correlated with total phosphorus concentrations and that DOC usage by lake bacteria depends strongly on the availability of phosphorus; the combination of carbon and phosphorus produced significantly greater bacteria concentrations than phosphorus alone. Similar results were found by Surbeck et al. [10], but the bacteria were FIB in a concrete-lined urban river where the source of phosphorus and organic carbon was not sediment, but the discharge from a wastewater treatment plant. Additionally, Jeng et al. [11] found that indicator organisms entered an estuary through stormwater runoff and settled to the bottom sediments, experiencing prolonged survival. Bolster et al. [12] were able to increase concentrations of E. coli in laboratory experiments of estuarine waters with high nutrient content.
Microcosm studies on FIB have been used by Craig et al. [13] and consisted of intact sediment cores taken from coastal areas with distinct sediment characteristics to determine the decay rates of E. coli in water and sediment. It was found that E. coli persisted in sediment, compared with overlying water. Small particle size and high organic carbon content were found to enhance survival; that is, sediments containing high proportions of clay and nutrients are more conducive to survival than sandy sediments.
2. Materials and Methods
2.1. Sample Collection
Water and sediment samples were collected during four events from an embayment (the flooded mouth of a tributary stream) contiguous with a manmade lake in northern Mississippi, USA. Water and sediment were collected from this water body and used in laboratory microcosms to simulate performance of BMPs with high FIB loadings.
The water and sediment samples were collected in Nalgene bottles (Nalgene Company, Rochester, NY) previously autoclaved at 121˚C in an EZ 9 autoclave (2340EA, Tuttnauer, Hauppauge, NY). Sediment samples were collected with 2-inch by 4-inch aluminum sediment core sleeves previously autoclaved at 134˚C. The sleeve was filled approximately halfway so that the sediment would represent the top 5 cm, which contains the most recent nutrient influx [14]. A sediment particle size analysis [15] was performed on the sediment and yielded size fractions of 71% sand and 29% silt.
Water samples were collected from the embayment surface, and sediment samples were collected from a submerged portion of the bank on four dates following precipitation events (Table 1). All samples were transported on ice, and microcosm set-up began once the samples were brought to the laboratory at the University of Mississippi.
2.2. Pre-Microcosm Preparation and Microbiological Analyses
The sediment was mixed to make a homogeneous soil matrix in order to provide each microcosm with similar sediment characteristics. This was accomplished by putting the sediment from the Nalgene bottles in an aluminum pan and manually mixing. Afterwards, sediment for two of the microcosms was sterilized by autoclaving at 121˚C.
Next, bacteria were extracted from the unsterilized sediment within three hours of collection. This was done following a procedure adapted from Craig et al. [16] and Jeong et al. [17]. Once the supernatant of the extraction procedure was retrieved, it was processed for analysis of total coliform and E. coli, in duplicate, using the defined-substrate method Colilert (IDEXX Laboratories, Inc., Westbrook, ME), in units of most probable number (MPN) per kg, and recorded. FIB analysis was performed on water samples using the Colilert method, but with concentration results in units of MPN per 100 ml (or 1
Table 1. Summary of initial concentrations of constituents in embayment water and sediment used for microcosm studies.
deciliter, dl). Colilert is known as Standard Method 9223 [18].
2.3. Microcosm Preparation
Following the sediment preparation and FIB quantification described above, seven microcosms were set up in 500-ml glass Erlenmeyer flasks. When necessary, water was filter-sterilized with a 0.22-micron filter (Model number 8532, Corning, Corning, NY). Sediments were sterilized by autoclaving. The microcosms were incubated in a shaker incubator (Classic C24, New Brunswick Scientific, Edison, NJ) rotating at 60 rpm to prevent settling of FIB, while mimicking the slow movement of water in a detention basin or constructed wetland. The shaker incubator was stopped for sample collection, but bacterial settling would not have occurred in the few minutes in which the shaker was stopped. An incubation temperature of 30˚C was used for all experiments as a representation of a summertime (i.e., recreation) water temperature. The incubator had a glass cover and was located near a window, exposing the microcosms to sunlight during the daytime.
Figure 1 is an illustration of Microcosms 1 through 5, which are described next. Microcosm 1 was composed of embayment water alone and was used to monitor FIB die-off free from the influence of bed sediment. Microcosm 2 was composed of embayment water and bottom sediment. This combination simulated natural conditions and was used to monitor FIB in the water as affected by nutrients associated with the water and nutrients and FIB