Decontamination efficacy of common liquid disinfectants against non-spore-forming biological agents in soil matrices

Abstract

Aims:

The purpose of this study was to evaluate decontamination efficacy, within three soil types, against Yersinia pestis, Burkholderia pseudomallei, and the Venezuelan Equine Encephalitis virus (VEEV).

Methods and Results:

One of three liquid disinfectants (dilute bleach, Virkon-S or Klozur One) was added to three soil types (sand, loam, or clay) and allowed contact for four pre-spike durations: 0, 15, 30 and 60 min. Y. pestis, B. pseudomallei, or VEEV was then spiked into the soil (10 microliters or approx. 1 × 10⁷ CFU or PFU into 1 g soil) and decontamination efficacy assessed at post-spike contact times of 10 or 60 min at ambient environmental conditions. Across all soil types, sandy soil resulted in the least quenching to all three disinfectants tested as shown by sustained decontamination efficacy across all pre-spike and post-spike timepoints. Clay and loam soil types exhibited quenching effects on the hypochlorite and peroxygen based disinfectants (dilute bleach and Virkon S) and in general resulted in decreased efficacy with increased pre-spike contact time. The sodium persulfate (Klozur One) performance was the most consistent across all soil types and pre-spike contact times, resulting in greater efficacy with increased post-spike time.

Conclusions:

Liquid disinfectants can provide high levels of decontamination in soil for both viral and non-spore-forming bacterial select agents. Hypochlorite and peroxygen based disinfectants used in soils containing higher organic content (loam or clay) may require extended contact times or re-application of liquid disinfectant, in as little as 15 min of application, to achieve a 6-log reduction.

Significance and Impact of the Study:

These results provide information for the performance of three disinfectants in soil against non-spore-forming select agents. These data may aid response decision makers following a biological contamination incident by informing the selection of disinfectant as well as the re-application time to achieve effective site remediation.

INTRODUCTION

Non-spore-forming biological agents have been shown to survive outside their host for days to months and are influenced by factors such as environmental conditions and the materials with which they have been applied (Calfee & Wendling, 2012; Eisen et al., 2008; Rogers et al., 2016; Sagripanti et al., 2010; Tong et al., 1996; Wood et al., 2015; Wood et al., 2018). Following a wide-area biological incident, many complex fomites including soil may become contaminated, requiring extensive site remediation.

Yersinia pestis is a highly infectious gram-negative, non-spore-forming, coccobacillus and is the causative agent of plague. Y. pestis is categorized by the Centers for Disease Control and Prevention (CDC) as a Tier 1 select agent due to its low infectious dose (as few as 10 organisms), virulence, natural availability, and its ability to spread via multiple modes of transmission including aerosol exposure (Ray & Ryan, 2004).

Venezuelan equine encephalitis virus (VEEV) is an enveloped single stranded RNA virus that can cause disease in both equids and humans. Human infection normally results in a self-limiting, incapacitating disease characterized by fever, headache, lymphopenia, myalgia, and malaise (Kumar et al., 2018). Additionally, severe neurological disease, which includes fatal encephalitis, can occur. Because of the ease of production, high infectivity, potential for aerosolization and disease state associated with infection, there is a concern for the viruses to be used as bioweapons and is therefore listed by CDC as a Select Agent.

Burkholderia pseudomallei is a gram-negative, non-spore-forming rod-shaped bacterium and is the causative agent of melioidosis. B. pseudomallei is readily isolated from wet soils in countries bordering the equator (Lafontaine et al., 2013). The CDC classifies B. pseudomallei as a Tier 1 select agent due to its availability, stability, antibiotic resistance and long incubation period making disease difficult to diagnose.

Decontaminating soil is challenging, whether it be following a wide-area biological incident or a small-scale contamination event resulting from wild animal infection. The ability to decontaminate biological pathogens in soil is highly dependent on the soil characteristics including organic content, chemical constituents, and physical properties such as density, particle sizes, and porosity (Wood & Adrion, 2019). Inactivation of contaminants buried within the soil column is especially challenging, as disinfectants must percolate down through the soil column prior to reaching the contaminants. This process may result in disinfectant chemistry being quenched and thereby reduce efficacy at depth. Understanding the quenching or neutralizing capacity of various soil types, against common disinfectant chemistries, is therefore important.

Most literature related to the decontamination of soil matrices have focused on spore-forming bacteria such as B. anthracis and liquid disinfection applications. Liquid disinfectants offer ease of use and application as well as the ability to penetrate the soil column. Liquid disinfectants such as 5% formaldehyde in seawater were shown to be effective, however, it required copious amounts and repeated applications of liquid disinfectants (Manchee et al., 1994). Other less effective liquid treatments include acidified chlorine bleach (Wood et al., 2011), peroxyacetic acid (U.S. Environmental Protection Agency, 2010), or aqueous chlorine dioxide (U.S. Environmental Protection Agency, 2012). Volumetric-applied disinfectants, such as fumigants (U.S. Environmental Protection Agency, 2013, 2017), have shown surface level efficacy but demonstrate challenges penetrating to lower depths (beyond 5 cm) within the soil column.

Several liquid disinfectants have previously shown effectiveness against various biological select agents on porous and non-porous building related surfaces (Calfee & Wendling, 2015; Richter et al., 2018; Rogers et al., 2005; Rogers et al., 2007; Rogers & Choi, 2008; Wood et al., 2016). The purpose of this evaluation was to study the effects of various non-sterilized soil types (clay, sand and loam) on the efficacy of three liquid disinfectants (dilute bleach, Virkon S and Klozur ONE) with three different active ingredients (hypochlorite, peroxygen, sodium persulfate, respectively), against Y. pestis, B. pseudomallei and VEEV. Non-sterilized soil was selected to ensure disinfectant quenching effects were representative of near field like conditions. Decontamination efficacy was determined for four pre-spike contact times (0, 15, 30 and 60 min) where the selected liquid disinfectant was allowed to dwell with the selected soil type, prior to inoculation of the microorganism, to simulate disinfectant-soil interactions that may occur in real-world applications, where disinfectant must percolate down through the soil column to reach contaminants at depth. In addition, decontamination efficacy was assessed for two post-spike contact times (10 and 60 min), where the select agent was allowed to dwell with the remaining (surviving quench) disinfectant within the soil. Decontamination efficacy data are presented as both total recovery and log reduction.

MATERIALS AND METHODS

Test organisms

Burkholderia pseudomallei strain K96243 was obtained from Northern Arizona University (NAU), prepared in lysogeny broth Lennox (LB Lennox, BD Cat. No. 240230) supplemented with 5% glycerol (Teknova Cat. No. G8797) (LBG), and stored at ≤−80° until used to initiate each experiment. A Gram-stain was performed on the B. pseudomallei stock and the colony morphology was confirmed to be consistent with published descriptions (Gilad et al., 2007). Fresh cultures were prepared in advance of each day that testing was performed by inoculating 100 μl of thawed stock material into 10 ml of LBG. The starter culture was incubated for 21 ± 3 h at 37 ± 2°C and 250 rev min⁻¹. The late log phase culture was diluted with fresh medium to an optical density at 600 nm (OD₆₀₀) of 0.2 ± 0.05 (SmartSpec 3000 Spectrophotometer, Bio-Rad), and 500 μl of diluted culture was used to inoculate an overnight culture of 100 ml LBG. The overnight culture was incubated for 19 ± 1 h at 37 ± 2°C and 250 rev min⁻¹. Following incubation, the overnight culture was diluted with fresh medium to an OD₆₀₀ of 2.1 ± 0.1. The final titre was determined by analysing 1:10 serial dilutions of the test organism suspension prepared using 10% Dulbecco's Phosphate Buffered Saline (DPBS, HyClOne Cat. No. SH30378.03) supplemented with 0.01% gelatin (Acros Cat. No. 61199) (BSC) and plated onto Ashdown's Selective Agar (Hardy Cat. No. G252). Plates were incubated at 37 ± 2°C for 72 ± 24 h and colony forming units (CFU) were enumerated.

VEEV subtype IAB was obtained from the University of Texas Medical Branch (UTMB) and prepared in routine cell culture medium, including minimum essential medium (MEM, Corning Cat. No 10-009-CV) and fetal bovine serum (FBS, ATCC Cat. No. 30-2020). The stock virus was propagated using Vero E6 cells (ATCC No. CRL-1586), harvested from the lysed cells with a target titre of approximately 10⁹ plaque-forming units (PFU) ml⁻¹, and stored at ≤−80°C. Titers were determined by analysing 1:10 serial stock dilutions in a plaque assay on Vero E6 cells, incubated for approximately 48 h, and the PFU enumerated.

Yersinia pestis strain CO92 was obtained from the University of Chicago (Chicago, IL), prepared in heart infusion broth (HIB, BD Cat. No. 238400) and stored at ≤−80°C until used to initiate each experiment. A Gram-stain was performed on the Y. pestis stock and the colony morphology was confirmed to be consistent with published descriptions (Josko, 2004). Fresh cultures were prepared in advance of each day that testing was performed by transferring several colonies from a streak plate (freshly grown) into HIB to obtain an OD₆₀₀ of 0.200 ± 0.05 and inoculating 500 μl of colony suspension into 100 ml fresh HIB. The overnight culture was incubated for approximately 24 h at 26 ± 2°C and 250 rev min⁻¹. Following incubation, the culture was diluted with fresh medium to an OD₆₀₀ of 2.56 ± 0.2 (approximately 2.05 × 10⁹ CFU/ml), then diluted 1:1 with fresh medium. The final titre was determined by analysing 1:10 serial dilutions of the culture using BSG and plated onto Yersinia Selective Agar (CIN, Hardy Cat. No. G220). Plates were incubated at 26 ± 2°C for 72 ± 24 h and CFU were enumerated.

Test materials

Three soil types were selected for testing (sand, clay and loam) and obtained from Agvise Laboratories. These soils were selected to represent a range of common soil types in the United States. To accurately assess the disinfectant quenching ability of each soil type, soils were not sterilized for testing. Selective media, further described below, were used in this study to sufficiently discriminate the target organisms from natural microorganisms contained within the soil. A full characterization of each soil type was performed by Agvise to assess density, moisture content, and nutrients (Table S1). Soil samples were prepared in capped 50 ml conical tubes (Thermo Scientific No. 339652) containing 1 ± 0.1 g of soil per tube.

Disinfectants

Three commercially available disinfectants were selected for efficacy testing against non-spore-forming biological agents. Virkon^™ S (pentapotassium bis(peroxymonosulphate) bis(sulphate)) was prepared by dissolving one tablet in 473 ml sterile water (HyClOne Cat. No. SH30529.03). Germicidal bleach (Clorox^®) was diluted 1:3 in sterile water to make dilute hypochlorite (2.1%). Klozur One (sodium persulfate) was prepared at a 0.5 M concentration in sterile water. Each disinfectant was prepared immediately prior to testing and used within 3 h of preparation.

Sample processing and data collection

All work was conducted in a biosafety level 3 (BSL-3) laboratory. For each soil type and disinfectant, three replicate soil samples were used for each pre- and post-spike contact time period. To conduct the pre-spike dwell, 350 μl of disinfectant was added to each soil sample tube. Quantification of the disinfectant volume was determined to be the amount required to thoroughly wet the soil sample without oversaturating and causing pooling. Tubes were agitated during the pre-spike dwell for dispersion of the disinfectant and saturation of the soil sample. After the prescribed pre-spike dwell time, samples were inoculated with approximately 1 × 10⁷ CFU or PFU per sample tube. A 10-μl aliquot of test organism (approximately 1 × 10⁹ CFU/ml or PFU/ml) was dispensed as one droplet on the surface of the soil sample, the tube was then agitated by hand to homogenize.

After test organism inoculation, samples remained at laboratory ambient conditions for 10 or 60 min, as outlined in Table 1. At the end of the post-spike contact time, decontamination activity was stopped by the addition of 3.86 ml of 2% sodium thiosulfate solution (STS, Sigma Cat. No. 217263), which was the calculated amount to achieve complete neutralization while not affecting viability of the bacteria or virus. This was followed by the addition of the appropriate extraction buffer (BSG for B. pseudomallei and Y. pestis, MEM + 5% FBS + 10% penicillin streptomycin [Gibco Cat. No. 15140-122, Carlsbad, CA] + 0.4% Gentamicin Solution [Sigma Cat. No. G1397] [virus diluent] for VEEV) to bring the final extract volume to 10 ml. Sample tubes were then vortexed at maximum speed for 2 min at room temperature. For each soil type and disinfectant pre- and post-spike contact time, three additional replicate samples were included as positive controls. Positive controls were pre-treated with 350 μl of extraction buffer, immediately inoculated, allowed the appropriate post-spike contact time, and extracted according to the procedure above. Additionally, one soil sample for each soil type was included as a blank (not inoculated) and included for each time point tested. The blank soil samples controlled for potential cross-contamination during testing and allowed identification of natural microorganisms within the soil.

TABLE 1.

Overview of experimental variables. Column headings indicate category of variables, row entries indicate the variables included for each category. Environmental conditions were not controlled during testing.

Organisms	Soil types	Decontaminants	Pre-spike (dwell) times (min)	Post-spike contact times (min)
Burkholderia pseudomallei K96243	Sand	Peroxygen (Virkon-S)	0, 15, 30, 60	10, 60
Yersinia pestis CO92	Clay	Sodium persulfate (Klozur One)
Venezuelan equine encephalitis virus	Loam	Dilute hypochlorite (dilute bleach)

To assess recovery from samples inoculated with B. pseudomallei or Y. pestis, resulting liquid extracts were 10-fold serially diluted in BSG. An aliquot (0.1 ml) of the selected dilutions and, when necessary, the undiluted extracts were plated onto agar plates in triplicate. The agar plates were incubated for 72 ± 24 h at 37 ± 2°C (B. pseudomallei) or 26 ± 2°C (Y. pestis). Colonies were enumerated to determine survivorship and reduction of the bacteria following exposure.

For samples inoculated with VEEV, resulting liquid extracts were 10-fold serially diluted in virus diluent. An aliquot (0.1 ml) of the selected dilutions and, when necessary, the undiluted extracts were plated onto 12-well monolayers of Vero E6 cells. The tissue culture plates were gently rocked every 15 min for 1 h at approximately 37° C to allow adsorption of the virus to the cells. The cultures were then overlaid with Eagle's Minimal Essential Medium (EMEM) that contained 10% FBS. The cultures were incubated for 48 ± 4 h at 37 ± 2°C and 5 ± 1% CO₂. Following incubation, crystal violet stain (Sigma Cat. No. HT901-8FOZ, St. Louis, MO) was added to the monolayers for 20 min, removed, and the cells rinsed with deionized water to remove the residual dye. Plaques were visualized as clearings in the purple monolayer of Vero E6 cells. The plaques were counted manually, and the number of PFU mL⁻¹ was determined by multiplying the mean number of plaques per well by the reciprocal of the dilution.

For B. pseudomallei, Y. pestis and VEEV assays, the theoretical limit of detection was 33 CFU or PFU when the undiluted sample was analysed. When endogenous flora required, 10-fold dilutions (10 ⁻¹ or 10⁻²) were used for analysis, adjusting the theoretical limit of detection to 333 or 3333 CFU or PFU, respectively.

RESULTS

A total of nine decontamination trials were conducted at ambient laboratory conditions, consisting of three tests per organism. In each trial, one disinfectant was tested in all three soil types, against one select agent.

Burkholderia pseudomallei

The mean number of bacteria applied to each soil sample across all tests was 7.23 ± 0.10 log₁₀ CFU. The mean recovery of B. pseudomallei for control samples of all soil types was 7.13 ± 0.13 log₁₀ CFU. Figure 1 shows the log reduction achieved at each pre-spike and post-spike contact time point, for each disinfectant tested. Across all soil types, efficacy was highest in sandy soil, for all three disinfectants tested. A 60-min contact time resulted in higher efficacy (log reduction) for each combination of disinfectant and pre-spike time.

FIGURE 1.

Efficacy of decontaminants against Burkholderia pseudomallei, represented by log reduction of bacteria, by soil type, pre-spike time and either 10-min (white) or 60-min (grey) post-spike contact time. Columns with stripes are samples that achieved complete kill. Theoretical limit of detection was 33 CFU.

Among the disinfectants tested, dilute hypochlorite demonstrated the highest efficacy against B. pseudomallei, producing the largest log reduction for each soil type. Pre-spike time did not appear to affect decontamination performance for clay and sand soil types at 60-min post-spike contact time, resulting in mean log reductions of 3.82 ± 0.12 and 6.83 ± 0.56, respectively. Efficacy decreased as pre-spike time increased for loam soil. Complete reduction of B. pseudomallei was achieved in sandy soil for 60-min post-spike contact times and 0-min, 15-min and 60-min pre-spike times.

The activated sodium persulfate performed similarly in clay and loam soils across all pre-spike durations, resulting in mean log reductions of 1.86 ± 0.13 and 2.26 ± 0.12, respectively, when allowed a 60-min post-spike contact time. Decontamination efficacy in sandy soil increased as pre-spike time increased, producing a maximum log reduction of 5.52 at 60 min pre-spike.

The peroxygen based disinfectant was least effective in decontaminating B. pseudomallei in clay and loam soil. For 60-min post-spike contact times, clay and loam soils had mean log reductions of 2.49 and 2.12, respectively, when samples were inoculated immediately following application of Virkon-S (0-min pre-spike time). All other pre-spike times resulted in a log reduction ≤0.2, with a B. pseudomallei recovery range of 54.6–110.6% recovery. Efficacy was higher for sandy soil but decreased as pre-spike time increased for both 10- and 60-min post-spike contact times. Complete reduction was achieved in sandy soil with a post-spike 60-min contact time and no pre-spike time.

Venezuelan equine encephalitis virus

The mean number/quantity of virus applied to each soil sample across all tests was 6.97 ± 0.03 log₁₀ PFU. The mean recovery of VEEV for control samples of all soil types was 6.84 ± 0.40 log₁₀ PFU. Figure 2 shows the log reduction achieved at each pre-spike and post-spike contact time point, for each disinfectant tested. In several cases, endogenous microorganisms within the soil samples impacted cell monolayers preventing the ability to identify viral plaques in the neat or 10⁻¹ dilutions. In these cases, the assay theoretical limit of detection was utilized to determine decontamination efficacy and is noted by symbols in Figure 2. Across all soil types, decontamination efficacy was highest in sandy soil for all three disinfectants tested, and a 60-min post-spike contact time generally resulted in higher efficacy than a 10-min post-spike contact time.

FIGURE 2.

Efficacy of decontaminants against Venezuelan equine encephalitis virus, represented by log reduction of virus, by soil type, pre-spike time, and either 10-min (white) or 60-min (grey) post-spike contact time. Theoretical limit of detection was 33 PFU when undiluted (10⁰) samples were analysed. LOD was adjusted to either 333 PFU (circle) or 3333 PFU (triangle), depending on the presence of endogenous flora, requiring enumeration of 10-fold diluted samples (10⁻¹ or 10⁻² respectively). Columns with stripes are samples that achieved complete kill.

The activated sodium persulfate demonstrated highest decontamination efficacy against VEEV, achieving significantly higher log reductions against the virus than the bacteria tested. Of the 60-min post-spike contact time samples, all but one sample achieved complete kill of virus, for an average log reduction of 6.49 across all soil types. The activated sodium persulfate also appeared to be less prone to quenching, as patterns of decreasing log reduction for increasing pre-spike time were not observed for any soil type.

Dilute hypochlorite performed similarly in clay and loam soils, with higher efficacy following a 60-min post-spike contact time (excluding a 0-min pre-spike time). Decontamination performance decreased as pre-spike time increased for both post-spike contact time points in both clay and loam soils. Sandy soil was more readily decontaminated by dilute hypochlorite than the other soil types, with all pre-spike timepoints resulting in comparable log reduction following a 60-min post-spike contact time. Decontamination performance also decreased for sandy soil as pre-spike time increased, when allowed a 10-min post-spike contact time.

The peroxygen based disinfectant was least effective in decontaminating clay and loam soil, achieving a similar pattern to that observed in B. pseudomallei with decreasing performance for increasing pre-spike time for both post-spike contact times tested. In sandy soil, complete kill of virus was achieved in all time points following the 0-min pre-spike for both 10 and 60-min post-spike contact times.

Yersinia pestis

The mean number/quantity of bacteria applied to each soil sample across all tests was 7.84 ± 0.77 log₁₀ CFU. The mean recovery of Y. pestis for control samples of all soil types was 6.76 ± 0.53 log₁₀ CFU. Figure 3 shows the log reduction achieved at each pre-spike and post-spike contact time point, for each disinfectant tested. In contrast to the trials for B. pseudomallei and VEEV, decontamination performance against Y. pestis was comparable across all three soil types.

FIGURE 3.

Efficacy of decontaminants against Yersinia pestis CO92, represented by log reduction of bacteria, by soil type, pre-spike time and either 10-min (white) or 60-min (grey) post-spike contact time. Theoretical limit of detection was 33 PFU when undiluted (10⁰) samples were analysed. LOD was adjusted to 333 PFU (circle) depending on the presence of endogenous flora, requiring enumeration of 10-fold diluted samples (10⁻¹). Columns with stripes are samples that achieved complete kill.

Dilute hypochlorite demonstrated the highest log reduction across all soil types, achieving complete kill in approximately 70% of time points tested. Complete kill was achieved in all 60-min post-spike contact time points. For the 10-min post-spike contact time points, the lowest log reduction was observed for the 30-min pre-spike time and all three soil types with an average log reduction of −0.007 ± 0.25 log₁₀. Clay and loam soils experienced decreasing efficacy for increased pre-spike time for the 0, 15 and 30-min pre-spike times. Complete decontamination was achieved for the 60-min pre-spike time for all three soils at a 10-min post-spike contact time. In this decontamination trial, endogenous microorganisms within the soil samples produced lawns in the neat dilution preventing the ability to identify Y. pestis colonies, therefore the theoretical limit of detection (333 CFU) was utilized to calculate efficacy.

The peroxygen based disinfectant was effective in Y. pestis reduction for all three soil types. For sandy soil, decontamination efficacy increased with pre-spike time for the 10-min post-spike contact time; highest efficacy was achieved with no pre-spike time with a 60-min post-spike contact time. Efficacy also increased with pre-spike time for clay soil at a 10-min post-spike contact time and highest log reduction with a 60-min post-spike contact time occurred at 15 min pre-spike. In loam soil, both post-spike contact times exhibited similar decontamination patterns with highest efficacy achieved at 15 min pre-spike, followed by decreasing performance with increasing pre-spike time.

The activated sodium persulfate was least effective in decontamination performance for all soil types. Highest efficacy for all soil types was achieved with a 15-min pre-spike and 60-min post-spike contact time, resulting in an average log reduction of 5.55 ± 0.78 log₁₀. All remaining time points achieved an average efficacy of 0.56 ± 0.72 log₁₀.

DISCUSSION

Outdoor environments may become contaminated with pathogens such as VEEV, Y. pestis, or B. pseudomallei due to an intentional release, epizootic contamination event, or environmental factors such as a seasonal monsoon resulting in contamination, or spread of contamination, within outdoor fomites such as soil (Ayyadurai et al., 2008; Baker et al., 2015). Many factors may affect decontamination efficacy of liquid disinfectants. Additionally, soil remains one of the most difficult matrices to decontaminate, due to its complex and varied constituents. Variables affecting decontamination efficacy include pathogen type, environmental conditions, disinfectant chemistry, soil type and depth of contamination within the soil column. The materials that microorganisms are in contact with are known to significantly impact the efficacy of antimicrobials (Calfee & Wendling, 2013; Wyrzykowska-Ceradini et al., 2019). The ability to rapidly remediate a contaminated area following a biological contamination incident requires not only an understanding of the disinfectants ability to achieve decontamination efficacy for a given pathogen type but also the potential quenching effect of various soil types on the selected disinfectant.

This study aimed to examine two main factors, first the ability to achieve decontamination efficacy on VEEV, Y. pestis, or B. pseudomallei in various soil types (Sand, Clay and Loam) selected to represent a range of low to high organic content 1.8, 3.9, and 4%, respectively. Secondly, this study examined the quenching effect of each soil type for the selected disinfectants, two of which are common hypochlorite or peroxygen based bacterial disinfectants (dilute bleach and Virkon S) and the third an activated sodium persulfate (Klozur ONE) typically used for in situ chemical oxidation of contaminants in soil and groundwater. Disinfectant quenching, most likely due to organics in the soil reacting with oxidant-based disinfectants, was quantified by reduction of efficacy in response to pre-addition of disinfectant to soil at four selected time-points for each pathogen and soil combination.

This evaluation utilized non-sterilized soil to ensure test methods were representative of near field like conditions. Various methods of soil sterilization (i.e. repeat autoclaving, e-beam or gamma irradiation) may change characteristics of the soil such as pH or organic content (Kelsey et al., 2010; McNamara et al., 2003). These changes may have an impact on the soils ability to neutralize various liquid disinfectants, thus raw unsterilized soil was utilized in this study. While using raw soil complicated quantification and in some cases limited the detection limit, due to endogenous microorganisms, the use of selective media (bacteria) and antibiotics (virus) allowed for sufficient resolution to demonstrate effects of soil on the selected liquid disinfectants.

The three liquid disinfectants evaluated resulted in varied levels of efficacy by soil type, organism, and pre- and post-spike contact time. The hypochlorite and per-oxygen based disinfectants (dilute bleach and Virkon-S) resulted in a time dependent quenching effect for sand, clay, and loam corresponding to the percent organic content of the soil (1.8, 3.9, and 4% respectively) as well as the pre-spike contact time with a noted exception of Y. pestis when exposed to dilute hypochlorite at the 60-min pre-spike exposure. This was an unexpected result since efficacy diminishing for the 10-min exposure samples but rebounded to achieve complete log reduction at the 60-min pre-spike exposure. This quenching effect was similar for both clay and loam soil types and was more pronounced at the shorter post-spike contact time of 10 min, suggesting that while neutralization is occurring, extended contact times may result in higher decontamination efficacy. The quenching effect resulted in decreased efficacy as pre-spike time increased for most samples between 15 and 60 min, suggesting that re-application of disinfectant after 1 h for hypochlorite-based disinfectants may also result in higher decontamination efficacy. Testing conducted in the sand soil type, which had the lowest percentage of organic material, resulted in minimal quenching and higher efficacy when compared to clay and loam.

The sodium persulfate-based product, on the other hand, did not exhibit the same quenching profile. This product is sold commercially to remediate soils contaminated with organic chemicals due to its ability to withstand soil quenching effects. This resulted in consistent decontamination efficacy, and in some cases, even increased as pre-spike time increased most likely the result of extended activation time with the soil. Sodium persulfate demonstrated best decontamination efficacy against VEEV, achieving greater than 6 log reduction after 60-min post-spike contact time in all soil types against the virus as compared to the bacteria tested. Conversely, both dilute hypochlorite and peroxygen demonstrated higher efficacy in all soil types for bacteria compared to sodium persulfate, apart from B. pseudomallei using Virkon-S.

While many factors may affect decontamination efficacy of non-spore-forming biological agents in soil, this study demonstrated the ability to characterize parameters affecting disinfectant quenching and subsequent impact to decontamination efficacy These data suggest that high efficacy against both non-spore-forming bacteria and virus is possible but may require extended contact times and multiple applications of disinfectant, depending upon soil type present and disinfectant chemistry selected. Further research is needed to better understand field applicability in terms of application rate of liquid disinfectant and subsequent penetration into compacted soil which may elucidate re-application time requirements to achieve ≥6 log reduction at a desired depth of potential contamination.

The data presented are intended to help guide disinfectant selection to aid remediation response following a biological contamination incident. Following a biological incident, these data will support rapid decision making to quickly conduct soil decontamination to reduce further exposure. Additionally, these data are valuable to regions of the world that battle chronic epizootic outbreaks and may facilitate reduction of frequent infections through strategic soil decontamination practices.