Inactivation of indicators and pathogens in cattle feedlot manures and compost as determined by molecular and culture assays

Authors

  • Marcus Klein,

    1. UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia
    Search for more papers by this author
  • Leearna Brown,

    1. UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia
    Search for more papers by this author
  • Nicholas J. Ashbolt,

    1. UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia
    2. National Exposure Research Laboratory, U.S. Environmental Protection Agency, Cincinnati, OH, USA
    Search for more papers by this author
  • Richard M. Stuetz,

    1. UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia
    Search for more papers by this author
  • David J. Roser

    1. UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia
    Search for more papers by this author

  • Editor: Julian Marchesi

Correspondence: Marcus Klein, UNSW Water Research Centre, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW 2052, Australia. Tel.: +61 2 9385 5102; fax: +61 2 9313 8624; e-mail: mklein@unsw.edu.au

Abstract

Accurate and conservative information about pathogen inactivation rates is needed as the basis for safe manure management on beef cattle feedlots. The survival of indicators and pathogens in faecal pen manure, stockpiled manure and manure compost was measured with autochthonous indicator bacteria (Escherichia coli, Clostridium perfringens, enterococci, total coliforms) and pathogens (Listeria monocytogenes, Campylobacter jejuni) using culture and/or real-time quantitative PCR (qPCR) methods. Additionally, the manures were incubated at 20, 37, 50 and 60 °C in microcosms to quantify the persistence of autochthonous microorganisms and selected process performance surrogates (Clostridium sporogenes, green fluorescent protein-labelled E. coli and L. monocytogenes). Estimated qPCR cell counts indicated that up to four orders of magnitude more target cells were present compared with the culturable counts. Corresponding T90 estimates were up to sixfold higher. This study demonstrates the benefits of nucleic acid-based quantification of pathogen inactivation in cattle manures and concludes that the concurrent analysis of microorganisms by molecular and culture methods provides complementary value.

Introduction

Cattle manure from feedlots is a valuable source of fertilizer, but zoonotic pathogens require control so they do not pose a risk to public or veterinary health and contamination of the natural environment (USEPA, 2005; Topp et al., 2009). At feedlots recontamination of cattle and potential human infections via environmental exposure to manure products are poorly understood. Consequently, the efficacy of manure control strategies resulting from cattle production needs to be further assessed. Manure composting and stockpiling are common practices, which are designed to reduce pathogen numbers for safe reuse (Larney & Hao, 2007; Martens & Bohm, 2009; Moral et al., 2009). Temperature is considered a major factor determining pathogen inactivation during storing and composting of manure, complemented by antagonistic microbial activity (Van Herk et al., 2004; Hutchison et al., 2005a; Nicholson et al., 2005; Larney & Hao, 2007). However, quantitative data on microbial inactivation rates and the influence of temperature in these materials are still limited, if not controversial (Inglis et al., 2010).

There are a number of reasons for undertaking new studies of microbial inactivation rates in animal products such as fresh and composted manure. Waste management and public health guidelines now need to be evidence-based, which may include the use of quantitative microbial risk assessment (QMRA) (Dixon, 2004; Topp et al., 2009). As a result, there have been a number of studies on the inactivation of autochthonous or inoculated zoonotic pathogens in cattle manures and the inactivation of pathogens after manure land application (Avery et al., 2004; Hutchison et al., 2004; Jiang et al., 2004; Hutchison et al., 2005b; Nicholson et al., 2005; Sinton et al., 2007). In addition, laboratory-based microcosms and small-scale in situ experiments have simulated the inactivation of native and inoculated pathogens or the fermentation of liquid manures (Lebuhn et al., 2003; Islam et al., 2004; Lemunier et al., 2005; Fremaux et al., 2008). These studies are in contrast to the previous practice that focused on faecal indicator bacteria, which were assumed to index pathogens of concern.

Stressful conditions present in many environmental materials facilitate the development of viable but nonculturable (VBNC) organisms (Lleo et al., 2005; Oliver, 2010). Pathogens or indicators in such a state lose their ability to grow in/on culture media, but may retain their infectivity. Therefore, this dormancy state can result in an underestimation of infectious pathogen numbers (Garrec et al., 2003; Lebuhn et al., 2004; Wery et al., 2006), increasing the uncertainty of data obtained via culture methods in QMRA (Haas et al., 1999).

Real-time quantitative PCR (qPCR) has been demonstrated as a reliable alternative technique for quantifying indicators and pathogens in a range of liquid and solid faecal materials of bovine origin including compost (Ibekwe & Grieve, 2003; Lebuhn et al., 2003; Inglis & Kalischuk, 2004; Cook & Britt, 2007; Klein et al., 2010b). The assay sensitivity with solid wastes is necessarily lower than for culture-based methods because of well-known constraints like DNA loss during isolation, the presence of PCR inhibitors and the limited sample size amplified (Zhang & Fang, 2006; Rapp, 2010).

The present work builds on a survey of in situ bacterial indicator and pathogen numbers in cattle manure (Klein et al., 2010a). It has reinvestigated the persistence of bacterial indicators and zoonotic pathogens in three major manure types produced at beef cattle feedlots in Australia using both culture-based and qPCR approaches.

Materials and methods

Experimental strategy

The inactivation of autochthonous and inoculated faecal indicators and pathogens was monitored in pen manure, stockpiled manure and manure compost on-site and/or using laboratory-based microcosms (Table 1). The microcosms provided temperature/humidity-controlled conditions to estimate the persistence of these microorganisms in all tested manures. All studies were performed over a period of 112 days except the microcosms with stockpiled and compost manure (69 days). Due to the low levels of target bacteria in stockpiled and compost manure, inactivation was analysed after inoculation with three structurally different model microorganisms. To quantify both culturable and total cell-specific counts (culturable, nonculturable, injured or dead cells), samples were analysed in parallel by culture techniques [most probable number (MPN), spread plating] and qPCR. In the case of compost/aged manure microcosms, two of the inoculated bacteria were chromosomally labelled with a functional single-copy gene for the green fluorescent protein (GFP) to make possible comparative quantification and the specificity of their detection in the presence of related autochthonous bacteria.

Table 1.   Inactivation analyses of autochthonous and inoculated target microorganisms in cattle feedlot manures
ExperimentManure typeDetection methodEscherichia coliEnterococciTotal coliformsEnterococcus faecalisClostridium perfringensClostridium sporogenesListeria monocytogenesCampylobacter jejuni
  • *

    Escherichia coli and Listeria monocytogenes chromosomally labelled with GFP gene.

Onsite surveyStockpileMPNxxx     
qPCRx  x    
CompostMPNxxx     
qPCRx  x    
Native microcosmPenMPNxxx     
qPCRxx xx xx
Inoculated microcosm*StockpilePlatingx    xx 
qPCRx     x 
CompostPlatingx    xx 
qPCRx     x 

Bacterial inactivation assessment in situ

All manures were collected from beef cattle feedlots in south-eastern Queensland, Australia (27°S, 151°E). Two representative aged manure types were sampled from two facilities with different management practices: ‘stockpiled manure’ was collected from a mound of freshly screened material harvested from cattle pens; ‘compost manure’ was taken from a compost windrow consisting of harvested pen manure mixed with 20% (w/w) sawdust. For in situ inactivation studies, duplicate samples were drawn after 0, 7, 14, 28, 56 and 112 days at a depth of 10–20 cm. Manure temperatures were recorded using temperature data loggers (Thermochron iButton, East Troy, WI). The field samples were stored at 4 °C in the dark, and processed within 24 h of collection. All samples were analysed twice for autochthonous faecal indicators by culture and qPCR methods at each time point (Table 1).

Bacterial inactivation assessment in microcosms

For the pen manure microcosm, approximately 1 kg of manure [45–55% moisture content (w/w)] was collected from three independent feedlot pens from a depth of 0–5 cm each, and composited. The thoroughly mixed material was divided into 50-g portions in sterile 70-mL polypropylene containers (Sarstedt, Nümbrecht, Germany) and incubated in duplicate at 20, 37 and 50 °C. The screw caps of the sample containers were slightly loosened to enable gas exchange. The water content was maintained at 51–57% (w/w) by keeping the aliquot containers in 3-L glass jars with water evaporation trays inside. Manure samples (1 g wet weight) were withdrawn after 0, 7, 14, 28, 56 and 112 days from the duplicate microcosms for analysis of autochthonous faecal indicators and pathogens by culture and/or qPCR methods (Table 1).

For the microcosms based on stockpiled and compost manures, early to medium-stage manures (c. 3 weeks) with moisture contents of 20–25% (w/w) were collected from operational manure stockpiles and compost windrows at a depth of 0–10 cm. To measure the bacterial inactivation inside the stockpiled and compost manures and also possible differences of bacterial cell wall structures on the bacterial survival, the representative but structurally different Escherichia coli, Listeria monocytogenes and Clostridium sporogenes were separately inoculated into microcosms (Table 1). Differences in cell wall structures should have a general impact on their survival. Escherichia coli BTF132 and L. monocytogenes BTF161 (both BTF, North Ryde, Australia) were used as models for Gram-negative and Gram-positive bacteria, respectively. Both possessed a single copy of the GFP gene stably integrated within their chromosome. They were cultured and harvested as reported earlier (Klein et al., 2010b). The Gram-positive and spore-forming C. sporogenes ATCC 19404 was used as a third model indicator for inactivation of environmentally persistent microorganisms. It was grown on Reinforced Clostridial agar (RCM) (Oxoid, Basingstoke, UK) within anaerobic jars at 35 °C for 48 h before harvesting colonies by scraping the agar plate surface and resuspending in phosphate-buffered saline (PBS). Bacterial concentration was estimated by OD600 nm measurement before inoculation. Two hundred-gram portions of the stockpiled manures and manure composts were inoculated with 107–108 bacteria g−1 manure (wet weight) and homogenized for 5 min at 200 min−1 using a mechanical stirrer (RW20, IKA, Staufen, Germany) to yield an even consistency and a moisture content of 20–25% (w/w). The inoculated manures were incubated in sterile 500-mL polypropylene containers (Sarstedt) as described for pen manure microcosms at 20, 37, 50 and 60 °C. Similarly treated microcosms without inocula were incubated in parallel as negative (E. coli and L. monocytogenes) or background (C. sporogenes) controls. The inoculated bacteria were enumerated by culture and qPCR methods 30 min postinoculation (t=0) and at indicated time points up to 69 days postinoculation by withdrawing triplicate 1-g samples from each microcosm (Table 1).

Quantification of indicators by culture methods

To recover the bacteria from the manure, 5 mL of PBS were added to manure samples followed by vortexing at highest speed for 10 s. Enumeration of autochthonous E. coli, enterococci and total coliforms was performed using the Colilert®-18 and the Enterolert® detection kits (IDEXX, Rydalmere, Australia), as described previously (Klein et al., 2010b). Inoculated GFP-labelled bacteria were enumerated in manure supernatants by spread-plating of decimal dilutions on to trypticase soy agar (Oxoid). RCM agar (Oxoid) was used for the enumeration of C. sporogenes. Background levels of native Clostridium spp. were determined in uninoculated samples and subtracted from the total counts. Bacterial concentrations were expressed as MPN g−1 or CFU g−1 manure (dry weight). The water contents of the samples were determined after drying at 105 °C for 24 h.

Quantification of indicators and pathogens by qPCR

The diluted manure supernatants containing target bacteria were centrifuged (9000 g, 15 min) and sequentially washed in PBS followed by 0.85% KCl to minimize the amount of possible extracellular bacterial DNA and PCR inhibitors. After isolation of total genomic DNA using a modified bead-beating procedure (Klein et al., 2010b), autochthonous bacterial indicators (E. coli, Enterococcus faecalis, Clostridium perfringens), autochthonous pathogens (L. monocytogenes, Campylobacter jejuni) and the inoculated GFP-labelled model bacteria were quantified by qPCR (Table 2). All PCR targets were single-copy genes. Forty qPCR cycles were run using SYBR® Green chemistry followed by data analysis as published earlier (Klein et al., 2010b).

Table 2.   Target sequences, cycling conditions and oligonucleotides for quantification of bacterial indicators and pathogens in cattle manures by qPCR
Target organismTarget gene; GenBank no.Annealing temperature; primer extension timeOligonucleotide names and references
Escherichia coliGlucuronidase; S6941460°C; 15 sEco-F, Eco-R (Shannon et al., 2007)
E. coli BTF132GFP; L29345.155°C; 15 sForward, Reverse (Klein et al., 2010b)
Clostridium perfringensAlpha toxin; AY27772460°C; 15 scpaF, cpaR (Gurjar et al., 2008)
Campylobacter jejuniVS1; X7160355°C; 20 sForward, Reverse (Stonnet & Guesdon, 1993)
Listeria monocytogenesListeriolysin; M2419960°C; 15 sForward, Reverse (Nogva et al., 2000)
L. monocytogenes BTF161GFP; L29345.155°C; 15 sForward, Reverse (Klein et al., 2010b)

Inactivation rates and statistics

Inactivation rates were calculated as T90 values and first-order reaction rate constant k, where Ct=C0× ekt (Ct is the concentration at time t, C0 the initial concentration and k the first-order inactivation rate constant, k=2.303/T90) (Burge et al., 1981). All data were assembled in an MS Excel table database and regression parameters were obtained using the LINEST array function. T90, k and correlation coefficients (R2) were calculated from the generated statistics. Where inactivation curves were judged to be biphasic, the statistics for both phases were calculated. Breaking time points between two phases were defined when the linear correlation of the adjoining curves was maximal. Half detection limit values were used in calculations where samples were below the detection limit. All SDs were calculated from logarithmic data values based on a minimum of three independent samples.

Results

Inactivation of indicators in stockpiled manure and manure compost in situ

The reduction of indicator numbers was monitored under in situ conditions in stockpiled manure and manure compost. Estimated T90 values ranged from 9.4 to 33 days based on MPN assays and from 14 to 62 days based on qPCR. Similar inactivation rates were obtained for E. coli and E. faecalis/enterococci (Table 3). Slightly lower T90 values were observed for all indicators in the stockpiled manure compared with manure compost. Inactivation rates estimated by qPCR were consistently higher than the equivalent culture-based values, indicating the greater persistence of target DNA derived from either unculturable or dead target cells. Overall, initial indicator concentrations were higher in stockpiled manure compared with manure compost by approximately two orders of magnitude. Autochthonous bacterial pathogens were detected sporadically by qPCR, but their levels were generally too close to the detection limit to reliably estimate inactivation rates (results not shown). The sample temperatures over the monitoring period for the specially established stockpiled manure and manure compost were 33–39 °C and 25–34 °C, respectively, potentially accounting for the more rapid inactivation in stockpiled manure.

Table 3.   Inactivation rates (T90; k) of autochthonous bacterial indicators in stockpiled cattle manure and compost windrows in situ
MicroorganismManure typeAssay methodInitial concentration (log10 g−1)nT90 (days)k (day−1)R2*
  • *

    Statistical significance P<0.01, P=0.01–0.05 (in italics) or P>0.05 (in parentheses).

  • Detection limits were five bacteria g−1 (MPN) and 600 copies g−1 (qPCR).

  • Results other than t=0 were less than the detection limit. Detection limit values were used for regression analysis.

  • §

    Specific for Enterococcus faecalis.

  • n, number of independent data values.

Escherichia coliCompostqPCR3.98<570.041(0.57)
MPN2.110280.082(0.71)
StockpileqPCR6.116140.160.73
MPN4.410100.220.92
EnterococciCompostqPCR§3.88<620.037(0.99)
MPN1.610330.0690.83
StockpileqPCR§4.816470.0490.85
MPN4.1149.40.25(0.58)
Total coliformsCompostMPN2.110220.100.89
StockpileMPN4.410170.140.91

Inactivation of indicators and pathogens in pen manure (microcosm)

The indicators showed a clear decrease in T90 with increasing temperature (Table 4). As seen with stockpiled manure and manure compost in situ (Table 3), the inactivation rates for enterococci and E. coli observed by culture were similar. Up to twofold higher T90 values were observed with E. coli and E. faecalis analysed by qPCR, indicating longer persistence of the corresponding genomic DNA, particularly at elevated temperatures. Very high or even negative T90 values were seen for total coliforms at 20 and 37 °C, indicating possible bacterial regrowth. More than 104C. perfringens copies g−1 were detected by qPCR. Inactivation rates estimated by qPCR were about twofold faster at 50 °C compared with 37 °C.

Table 4.   Inactivation rates (T90; k) of autochthonous bacterial indicators and pathogens in microcosms based on pen manure
MicroorganismAssay methodInitial concentration (log10 g−1)Temperature (°C)nT90 (days)k (day−1)R2*
  • *

    Statistical significance P<0.01, P=0.01–0.05 (in italics) or P>0.05 (in parentheses).

  • Detection limits were five bacteria g−1 (MPN) and 600 copies g−1 (qPCR).

  • Possible regrowth.

  • §Results other than t=0 were less than the detection limit. Half detection limit values were used for regression analysis.

  • n, number of independent data values.

Escherichia coliqPCR4.7209530.0430.72
3710290.0790.94
506140.17(0.22)
MPN3.72012350.0660.84
379150.160.77
5087.70.300.80
Enterococcus faecalisqPCR5.82010610.0380.70
3710380.0600.95
5010290.0790.97
EnterococciMPN3.72010440.0520.75
378160.150.75
5068.30.280.75
Total coliformsMPN3.8208  
379  
5087.90.290.78
Clostridium perfringensqPCR4.52010340.0680.53
379320.0710.70
506150.150.76

Some of the more common autochthonous pathogens like L. monocytogenes and C. jejuni were also detected in the pen manure at 2.0 × 103 and 5.0 × 104 genome copies g−1, respectively. qPCR detectable C. jejuni DNA was below the detection limit at all incubation temperatures after 7 days, resulting in T90 values under 3.6 days (k>0.64). For L. monocytogenes DNA, calculation of inactivation rates was impaired by marginally low starting concentrations and limited statistical significance.

Inactivation of inoculated indicator bacteria in stockpiled manure and manure compost (microcosm)

In the experiments with inoculated model bacteria, the most rapid decline in numbers of culturable bacteria was seen with E. coli (T90=<0.3–4.4 days), followed by the Gram-positive L. monocytogenes (T90=0.51–126 days) and the spore-forming C. sporogenes (T90=1.9–185 days) (Fig. 1, Tables 5 and 6). Escherichia coli was rapidly inactivated in both manure types if exposed to 37 °C and higher. In contrast to E. coli and C. sporogenes, the recovery of culturable L. monocytogenes from both manures showed a distinctive biphasic inactivation pattern, particularly at 37 and 50 °C. Slightly faster inactivation of L. monocytogenes was observed in the manure compost microcosm by culture methods at these temperatures. No major differences in inactivation were seen between stockpiled manure and manure compost for C. sporogenes, which appeared to be the most persistent indicator. For each time point, background levels of autochthonous Clostridium spp. detected in negative controls were at least two orders of magnitude lower than the concentration of all detected clostridia (results not shown). No significant reduction of C. sporogenes numbers was observed in the microcosms at 20 °C (Fig. 1). This indicates that their technical recovery was more or less unimpaired over the total experimental period.

Figure 1.

 Inactivation of Escherichia coli (▴), Listeria monocytogenes (+) and Clostridium sporogenes (◊) after inoculation into cattle manure microcosms containing stockpiled (a) or composted manure (b). Bacteria were quantified at the indicated incubation periods in independent triplicate samples by spread-plating. Error bars are ±1 SD. Detection limits were 50 bacteria g−1. Last detection value (69 days) not shown.

Table 5.   Inactivation rates (T90; k) of inoculated Escherichia coli, Listeria monocytogenes and Clostridium sporogenes in microcosms with manure from a stockpile
MicroorganismAssay methodTemperature (°C)nT90 (days)k (day−1)R2*
  • *

    Statistical significance P<0.01 or P=0.01–0.05 (in italics).

  • Results other than t=0 were less than the detection limit. Half detection limit values used for regression analysis.

  • 1st phase 0–6 days; 2nd phase ≥6 days.

  • §

    1st phase 0–2 days; 2nd phase ≥2 days.

  • n, number of independent data values.

E. coliqPCR2021120.190.98
3798.00.290.85
5091.61.410.88
Culture20184.30.520.96
376<0.3>7.80.99
506<0.3>7.80.99
L. monocytogenesqPCR2021360.0870.94
37155.60.540.95
50122.31.00.94
606<2.3>1.00.93
Culture2021110.210.84
379; 151.9; 1261.2; 0.020.69; 0.26
509; 151.6; 161.4; 0.140.73; 0.98
606; 120.52; 9.6§4.4; 0.240.99; 0.96
C. sporogenesCulture20151000.0230.71
3715460.0500.89
50155.80.400.90
60122.11.20.98
Table 6.   Inactivation rates (T90; k) of inoculated Escherichia coli, Listeria monocytogenes and Clostridium sporogenes into microcosms with compost manure
MicroorganismAssay methodTemperature (°C)nT90 (days)k (day−1)R2*
  • *

    Statistical significance P<0.01, P=0.01–0.05 (in italics) or P>0.05 (in parentheses).

  • Results other than t=0 were less than the detection limit. Half detection limit values used for regression analysis.

  • 1st phase 0–6 days; 2nd phase ≥6 days.

  • §

    1st phase 0–2 days; 2nd phase ≥2 days.

  • n, number of independent data values.

E. coliqPCR2021270.0860.85
37156.50.350.90
5091.71.40.99
Culture20184.40.530.97
376<0.9>2.60.99
506<0.3>7.80.99
L. monocytogenesqPCR2021650.0350.61
37187.40.310.93
50122.50.930.94
6063.20.720.98
Culture2021170.140.73
379; 151.6; 1211.4; 0.0190.98; (0.26)
509; 150.60; 8.24.1; 0.280.98; 0.86
606; 120.51; 14§4.5; 0.170.99; 0.76
C. sporogenesCulture20151850.0120.39
3715470.0490.76
5015100.220.97
60121.90.960.94

To determine the persistence of DNA corresponding to total culturable plus nonculturable E. coli and L. monocytogenes cells, the single-copy GFP gene within their chromosome was quantified by qPCR. Compared with the plating results, qPCR showed a markedly slower decline compared with the corresponding bacterial numbers in both stockpiled manure and manure compost microcosms (Fig. 2, Tables 5 and 6). This indicated that the majority of genome copies originated from nonculturable, injured or dead cells, particularly at lower incubation temperatures. Escherichia coli DNA was still detectable after 29 days at 37 °C, whereas the numbers for culturable E. coli dropped below the detection limit within 2 days after inoculation.

Figure 2.

 Decrease in genomic DNA copy numbers of Escherichia coli (▴) and Listeria monocytogenes (+) after inoculation into cattle manure microcosms containing stockpiled (a) or composted manures (b). Genome copy numbers were quantified by qPCR at the indicated incubation periods in independent triplicate samples also used for analysis by culture. Error bars are ±1 SD. Detection limits were 600 (stockpiled manure) and 1200 (compost manure) bacteria g−1. Last detection value (69 days) not shown.

In contrast to the results obtained by spread-plating, slower reduction of corresponding L. monocytogenes DNA was seen in both the compost and stockpiled manure, resulting in higher T90 values at all temperatures (Tables 5 and 6). After calculating the ratio between culture-detected bacteria and corresponding qPCR-detected DNA, excesses of more than four orders of magnitude of genomic DNA were estimated. The highest ratios were observed in the microcosms with L. monocytogenes in manure compost (not shown). In general, the reduction curves obtained after qPCR analysis showed a log linear response consistent with first-order inactivation kinetics (Fig. 2).

Discussion

The current study was designed to measure the inactivation of faecal indicators and pathogens in three types of feedlot cattle manures reflecting local management practice. Bacterial inactivation rates were measured using culture methods to quantify the decrease in culturable cells, whereas qPCR was applied to examine the reduction of specific DNA from intact viable and VBNC cells. Inoculated microcosms were used where the concentration of autochthonous microorganisms was low or below the detection limits.

Comparison of bacterial inactivation rates in stockpiled and compost manure material in situ was hampered by differences in starting concentrations of the target organisms, as stockpiled manure levels were generally two orders of magnitude greater than compost manure levels (Table 3). Indicator concentrations near the detection limit were present in the compost manure in contrast to the stockpiled manure, despite the fact that both processes were initiated with a comparable management level and manure age. One possible explanation could be that more favourable conditions existed for bacterial reduction during earlier manure management steps on the feedlot that is stockpiling aged manures. Nevertheless, most of the analyses by culture methods resulted in T90 values in good agreement with previously published data for related manure materials (Table 7). The corresponding qPCR results showed 1–2 orders of magnitude higher starting numbers, indicating the presence of intact bacteria that have lost the ability to become detected by culture methods (e.g. VBNC). However, the statistical value of the calculated inactivation rates was limited because of the low target concentrations leading to reduced numbers of samples above the detection limit for qPCR.

Table 7.   Comparison of inactivation rates of faecal indicators and pathogens in cattle manures determined by culture methods and qPCR
MicroorganismT90 (days)Manure typeTemperature (°C)Reference
Culture detection
 E. coli38Pen manure0–9Avery et al., 2004
46–38Pats on pasture−2–47Sinton et al., 2007
35–7.7Pen manure20–50This study
4.4–0.3Manure compost20–50This study
4.3–0.3Manure stockpile20–50This study
 Enterococci77–38Pats on pasture−2–47Sinton et al., 2007
44–8.3Pen manure20–50This study
 C. jejuni1.5Manure stockpile4–15Hutchison et al., 2005a
1.3–4Manure-amended soil10Ross & Donnison, 2006
 L. monocytogenes3.7Manure stockpile4–15Hutchison et al., 2005a
5.0Manure-amended soil5–21Jiang et al., 2004
17–0.5Manure compost20–60This study
11–0.5Manure stockpile20–60This study
 C. sporogenes160Manure-amended soil12–25Girardin et al., 2005
185–1.9Manure compost20–60This study
100–2.1Manure stockpile20–60This study
qPCR detection
 E. coli53–14Pen manure20–50This study
12–1.6Manure stockpile20–50This study
27–1.7Manure compost20–50This study
 E. faecalis61–29Pen manure20–50This study
 L. monocytogenes36–2.3Manure stockpile20–60This study
65-2.5Manure compost20–60This study
 C. perfringens34–15Pen manure20–50This study
 C. jejuni<3.6Pen manure20–50This study

To analyse the degradation of indicators and pathogens in more detail and over longer time periods to produce statistically robust results, laboratory-based microcosms were set up for all tested manures. Inactivation of microorganisms was particularly important in pen manure because this material is both ubiquitous on a quantity basis and it still may contain high pathogen numbers (Hutchison et al., 2005a; Klein et al., 2010a, b). Due to its exposure to wind and water run-off (USEPA, 2005; Topp et al., 2009), this information was considered of particular significance. As expected, the pen manure showed significantly higher starting concentrations compared with the older stockpiled and compost manures. This allowed for a precise statistical analysis of inactivation rates by both culture and qPCR methods using autochthonous microorganisms, even at 50 °C, when rapid die-off kinetics were observed. In particular, the inactivation of E. coli and enterococci at 20–37 °C corresponded closely to the T90 values seen in stockpiled and compost manures in situ, of similar ambient temperatures. In contrast, total coliforms in the pen manure microcosm revealed negative inactivation rates due to bacterial regrowth, most likely as a consequence of the higher content of nutrients and humidity (Garrec et al., 2003). The presence of low concentrations of autochthonous L. monocytogenes, possibly caused by high sample-to-sample variations, hampered the precise calculation of their inactivation in the pen manure. Campylobacter jejuni, initially present in high numbers, rapidly reduced in detectable DNA at all incubation temperatures, and inactivation rates were in good agreement with previous reports (Inglis et al., 2010; Klein et al., 2010a). The fast disappearance of high amounts of specific C. jejuni DNA within a relatively short time span of 7 days confirmed a high recycling rate of extracellular nucleic acids in manures of bovine origin (Lebuhn et al., 2004; Lleo et al., 2005). Hence, the nondetection of possible extracellular DNA from freshly disintegrated cells underlines the general value of the applied sample preparation/qPCR detection system for the quantification of entire bacterial cells.

Three common indicators and pathogens were used for inoculating microcosms based on stockpiled and compost manure to represent a spectrum of Gram-positive, Gram-negative and spore-forming bacterial targets. The use of GFP-labelled model strains in these microcosms minimized the possibility of inactivation patterns being a result of mixed biotypes with different persistence (Islam et al., 2004; Jiang et al., 2004; Fremaux et al., 2008). We further extended their utility by concurrently enumerating the inoculated GFP-labelled bacteria by both culture and qPCR assays. Autoclaving or other methods to eliminate native background levels was unnecessary, and so the autochthonous microbiota induced in the natural decomposition of the manures was disturbed minimally. Compared with the inactivation experiments based on autochthonous bacteria in pen manure, shorter T90 periods were detected for E. coli and L. monocytogenes at all incubation temperatures. One possible reason for this observation could be the suboptimal conditions in the older and dryer manures, but also less chance of bacterial regrowth. In general, inactivation patterns and rates for culturable organisms were comparable in magnitude with estimates from previous studies, particularly in the case of L. monocytogenes and C. sporogenes (Table 7). This, and the fact that mostly high correlation coefficients and statistical significance were observed, indicates that our experimental microcosms were credible systems for estimating inactivation rates under well-controlled conditions. Higher initial and biphasic inactivation rates were seen with culture-based assays involving L. monocytogenes (Fig. 1), possibly reflecting adaptations to altered biological stress factors or changing inactivation mechanisms. Nonlinear inactivation curves followed by tailing were previously observed under similar conditions (Hutchison et al., 2004; Hutchison et al., 2005b); however, they may also be explained by bacterial regrowth (Garrec et al., 2003) or the coexistence of heterogeneous bacterial populations with differing resistance to environmental stress (Jiang et al., 2003).

Measurement of the inactivation of inoculated C. sporogenes allowed for comparisons of inactivation rates for persistent spore formers with nonsporulating bacteria like E. coli or Listeria. Although their inactivation was slower, composting appears to be a credible means for inactivating other hardy pathogens, like for example Cryptosporidium, Giardia, Mycobacterium or Coxiella. Limitations to note were that the inoculated C. sporogenes were a model culture strain rather than an actual feedlot resident, and a species-specific molecular detection method for C. sporogenes was unavailable.

No major differences were seen between the compost and stockpiled manure microcosms, suggesting a comparable performance of both manure systems under the selected conditions. The fact that there were no obvious differences underlines the value of temperature as a strong inactivating factor (Van Herk et al., 2004; Hutchison et al., 2005a; Nicholson et al., 2005; Larney & Hao, 2007). In fact, the detected temperatures of the analysed stockpiled manure in situ were 5–7 °C higher on average, probably a consequence of their larger size compared with the smaller compost windrows. This indicated that unamended manure in large manure stockpiles might have sufficiently high residual organic carbon to achieve similar or even higher temperatures than the carbon-enriched compost.

The VBNC phenomenon has received considerable attention in environmental microbiology over the last decade (Oliver, 2010). The potential for pathogens in manure to enter such a state of inactivity has not been considered explicitly in the estimation of inactivation rates. However, this issue has been identified as a concern during the consideration of comparable environmental materials (Lebuhn et al., 2003, 2004; Artz et al., 2006; Wery et al., 2006). As opposed to the underestimation of potentially infectious organisms using culture methods alone, quantification by qPCR is more likely to lead to an overestimation of bacterial numbers due to detection of coisolated dead, but still unbroken cells containing genomic DNA. Therefore, by detecting the whole range of physiological cellular stages of such bacterial targets, covering all states from infectious to dead, qPCR appears to account for the uncertainty when assessing pathogen risks.

As a major result of the present study, we demonstrated using qPCR that genomic DNA of reversibly damaged or otherwise nonculturable pathogens can exceed the corresponding culturable bacteria estimates by more than four orders of magnitude, highlighting the potential of such environments to promote the accumulation of bacteria resting in a VBNC or otherwise inactive state. Consequently, qPCR-based inactivation rates were generally lower than their culture-based equivalents. The probability to detect significant amounts of extracellular nucleic acids was minimized using an optimized method for isolating intracellular DNA targets (Klein et al., 2010b). All inactivation data obtained by qPCR followed first-order kinetics without showing any bi- or multiphasic components, accentuating this method as a precise tool to monitor bacterial inactivation kinetics. Detection limits down to approximately 103 genome equivalent targets per gram manure, low SD values and high reproducibility were achieved, particularly where manures with moderate to high initial concentrations of microorganisms were analysed. However, further research remains to be performed to pinpoint the quantitative detection of exclusively viable and potentially infective pathogens.

In summary, the major outcomes of this study were as follows: (1) The studies provided precise pathogen/indicator inactivation rate estimates and variances for temperatures between 20 and 60 °C for stockpiled and compost manure. (2) We demonstrated the practicality of qPCR for quantifying the disappearance of ‘DNA from’ microorganisms in cattle manures that may outnumber culturable cells by several orders of magnitude. (3) The persistence of genomic DNA analysed by qPCR performed approximate first-order removal kinetics, indicating such modelling as being credible; (4) GFP-labelled indicator bacteria or pathogens inoculated into microcosms are a precise tool for studying VBNC and related effects in environmental samples. (5) The qPCR data supported the concern that current T90 estimates might underestimate pathogen survival and may be useful for determining more conservative numbers required in QMRA; For the future, we suggest that the concurrent estimation of culturable and total pathogen numbers should improve the reliability of QMRA application to livestock manure and promote renewed investigation of how the VBNC phenomenon affects pathogen risk estimates.

Acknowledgements

This work was supported by Meat and Livestock Australia (MLA). We would like to express our thanks to Robyn W. Tucker from Feedlot Services Australia Pty Ltd (FSA Consulting) for her contribution in fieldwork. The views expressed in this article are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency.

Ancillary