Use of Nitrogen‐15‐Enriched Escherichia coli as a Bacterial Tracer in Karst Aquifers

Abstract

Karst aquifers are susceptible to contamination by microorganisms, but relatively few studies have used bacteria as tracers. We demonstrate the utility of Escherichia coli enriched in the stable isotope nitrogen-15 ( N) as a novel bacterial tracer. Nonpathogenic E. coli from two springs in central Kentucky were grown on N-enriched media. Survival of E. coli and persistence of the isotopic signal were assessed in two sets of laboratory experiments conducted with sterilized spring water in dark microcosms at 14 °C. First, isotopically labeled bacteria survived for 130 d at concentrations within one log unit of the average initial value, and there was no significant difference in δ N values from Day 1 to Day 130. Second, water samples with E. coli were inoculated with either of two different species of protozoa (Tetrahymena pyriformis or Colpoda steinii). During 7 d, δ N values increased in T. pyriformis while bacterial populations decreased. In a field test, following a 2.1-cm rainfall, N-labeled E. coli, solutes (rhodamine WT dye and bromide), and latex microspheres were injected into a sinkhole approximately 530 m upgradient of a spring. Breakthrough of all tracers coincided, but microspheres were remobilized by subsequent storms, unlike other tracers. Enriched E. coli exhibited more tailing than solute tracers during the initial storm-flow recession. These results indicate that N-enriched E. coli is a viable tracer of bacterial transport in karst aquifers, although predation may attenuate the isotopic signal in systems that are not rapidly flushed.

Use of Molecular Markers to Compare Escherichia coli Transport with Traditional Groundwater Tracers in Epikarst

“…However, results differ from a 30‐m trace in saturated Swiss epikarst in which particulates preceded solute tracers (Goldscheider et al, 2008). Solute and particulate arrival times differed, as shown by initial BTCs, yet peaks overlapped during periods of high discharge, similar to results of Göppert and Goldscheider (2008) and Ward et al (2016). Microspheres arrived at the sampling location at the same time as LAI but reached maximum values >200 h later, in conjunction with HAI maxima during storm flow.…”

“…Particulate tracers such as DNA, wild‐type Escherichia coli ( E. coli ), Ralstonia eutropha bacteria, 1‐μm polystyrene beads (microspheres), and clay loam suspensions have been shown to arrive and/or peak in concentration prior to solutes such as fluorescent dyes, I − , pesticides, and NO 3 − within various karst, soil, and fractured crystalline rock settings (Atkinson et al, 1973; Champ and Schroeter, 1988; Orth et al, 1995; Ting et al, 2005; Goldscheider et al, 2008; Sinreich et al, 2009; Aquilanti et al, 2013; Firouzi et al, 2015). However, in other studies within karst aquifers, particulate tracers have peaked simultaneously with or after solute tracers (Ting et al, 2005; Goeppert and Goldscheider, 2011; Ward et al, 2016). Breakthrough curves (BTCs) can also show differing behaviors between tracers, as seen in an epikarst trace in which a solute had a broader BTC than bacteriophages (Flynn and Sinreich, 2010).…”

“…Strain‐level diversity of E. coli is very high, likely contributing to survival of the organism in environmental habitats (Foppen et al, 2010; Cook et al, 2011). Survival times listed in the literature for groundwater, surface water, and soils range from 8 to 12 h to several months (Personné et al, 1998; Wcisło and Chróst, 2000; Banning et al, 2002; Craig et al, 2004; Davis et al, 2005; Franz et al, 2011; Ma et al, 2014; Ward et al, 2016). Mortality of E. coli in water sources is dominated by predation and water chemistry (Wcisło and Chróst, 2000; Davis et al, 2005; Cook and Bolster, 2007).…”

Multiple-tracer tests for contaminant transport process identification in saturated municipal solid waste

“…Ward et al (1998) found that out of the available 'solute-type' tracers for aquifers and soils, the halogens are most suitable, of which bromide is often preferred (as it often has lower and more stable background levels). However, Stegemann et al (2006) showed non-linear bromide sorption to waste.…”

“…This is important since to achieve objective of the paper required a meaningful comparison of these estimates. This is fortunate since an inability to obtain a reliable estimate of cb t is common for field tests (Ward et al, 1998;Mathius et al, 2006;Hadermann & Heer, 1996;Haggerty et al, 2000). The difference between the field tests and these well-controlled column tests is that for the latter the flux rate and total porosity are accurately available thus providing better constraint.…”