survival of pathogenic bacteria during mesophilic anaerobic digestion of animal waste

Journal of Applied Bacteriology 1993, 75, 215-219

Survival of pathogenic bacteria during mesophilic anaerobic digestion of animal waste

Theresa E. Kearney, M.J. Larkin, J.P. Frost’ and P.N. Le~et t ’ ’~ School of Biology and Biochemistry, The Queen ’s University of Belfast, ’Agriculture Research Institute of Northern Ireland, Hillsborough I Co. Down and 2Department of Biological and Medical Sciences, University of Ulster, Coleraine, Northern Ireland, UK

4347/10/92: accepted 2 April 1993

T.E. KEARNEY, M.J. LARKIN, J.P. FROST AND P.N. LEVETT. 1993. T h e survival of pathogenic bacteria was investigated during the operation of a full-scale anaerobic digester which was fed daily and operated at 28°C. The digester had a mean hydraulic retention time of 24 d. The viable numbers of Escherichia coli, Salmonella typhimurium, Yersinia enterocolitica, Listeria monocytogenes and Campylobacter jejuni were reduced during mesophilic anaerobic digestion. Escherichia coli had the smallest mean viable numbers at each stage of the digestion process. Its mean T,, value was 76.9 d. Yersinia enterocolitica was the least resistant to the anaerobic digester environment; its mean T,, value was 18.2 d. Campylobacterjejuni was the most resistant bacterium; its mean T,, value was 438.6 d. Regression analysis showed that there were n o direct relationships between the slurry input and performance of the digester and the decline of pathogen numbers during the 140 d experimental period.

INTRODUCTION

The composition of material fed to anaerobic digesters is a major factor determining the growth and metabolic activity of both non-methanogenic and methanogenic bacteria and hence the production of methane during anaerobic digestion. The source of the feeding material and its composition are also important in determining the number of viable pathogenic bacteria present in slurry. I t has been reported that pig, cattle and poultry manures from inten- sive farming systems harbour pathogenic bacteria, inciud- ing Salmonella spp., Escherichia coli, Listeria monocytogenes and Campylobacter spp. (Jones and Matthews 1975; Jones 1980; Larsen and Munch 1982; Skirrow 1990). Farm slurry treatment would be advantageous to farmers if both the organic content of slurry and the numbers of pathogenic bacteria present were reduced during the process. This would reduce the time between spreading slurry on pasture and allowing animals to graze the pasture. Meso- philic anaerobic digestion is an effective process in reducing the organic content of slurry and consequently its biological oxygen demand. It has also been reported that mesophilic anaerobic digestion is effective in reducing the numbers of

Present address: Universrty of the West Indies, Faculty of Medical Sciences, Queen Elizabeth Hospital, Barbados. Correspondence t o : Dr T.E. Kearney, School of Biology and Biochemistry, Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, Northern Ireland, 1JK.

pathogenic bacteria, although the extent to which it does so depends upon the pathogen species and slurry source (Carrington et al. 1982; Turner et al. 1983; Gadre et at. 1986; Olsen and Larsen 1987). The change in bacterial viable numbers during anaerobic digestion may be evalu- ated using the decimation reduction time (T9,). This is defined as the time it takes for the viable counts of a population to decrease by one logarithmic unit (log,,) which is equivalent to a 90% reduction (Schlundt 1984). For example, within pig-fed digesters, Olsen and Larsen (1987) reported that Salmonella typhimurium had a T90 value of 2.0 d, whilst the T,, value for the same organism was 2.9 d in digesters fed with cattle slurry. Olsen and Larsen (1987) have reported that laboratory strains of Salm. typhimurium and E. coli have T9, values of 2.1 and 1.7 d respectively within a pilot anaerobic digester and, within a full-scale digester (100 m3), the T90 values of these bacteria were not significantly different. It has not been well documented how the indigenous pathogen populations, as opposed to laboratory attenuated strains, survive during anaerobic digestion of farm waste.

The aim of this study was to assess the effect of full-scale semi-continuous anaerobic digestion on the indigenous viable populations of E. coli, Salm. typhimurium, Yersinia enterocolitica, L. monocytogenes and Campylobacter jejuni. In addition, changes in the viable populations of these bacteria after separating the liquid and solid fractions of the digester effluent were examined.

216 T H E R E S A E . K E A R N E Y E T A L .

MATERIALS AND METHODS

Full-scale anaerobic digester

4 continuously stirred anaerobic digester, model F.G.B. !I0 (Farm Gas Ltd., Bishops Castle, Shropshire, UK), ivith a working volume of 210 m 3 was used to treat farm ivaste. T h e digester temperature was maintained at 28°C ,during anaerobic digestion by hot water heat exchangers. The water was heated by both combustion of the biogas m d electricity. Excess gas was piped into a gas storage bag 150 m3) situated beside the digester. T h e retention time of

-he slurry in the digester was 25 d. T h e quantity of slurry <ed to the digester and the removal of digester effluent were monitored and controlled during the anaerobic digestion process.

T h e digester was batch-fed daily but the effluent was continually pumped from the digester through pipes to a qelt press slurry separator (Farm Gas I d ) . T h e separated iquid fraction was piped into a lagoon and the remaining

tolid fraction was collected and subsequently composted in .i thermophilic aerobic composter unit.

Dlgester performance

The performance of the full-scale anaerobic digester was nonitored weekly during the 140 d experimental period by :he Department of Agriculture for Northern Ireland and :he Agriculture Development Advisory Service as part of a I year survey. Biogas production was mcasurcd weekly with :wo domestic-type gas meters (Schlumbuger Industries, Royton, Oldham, UK). T h e composition of the biogas was .malysed on site each week for CO, and H,S using gas rnalysis tubes (Gastec Corporation, Yokohama, Japan). The percentage methane present in each biogas samplc was 2alculated as :

The calorific value of biogas was measured direct11 by burning a gas sample (100 ml) in a calorimeter (Sigma). The water vapour content of the biogas was calculated to DC 14%) with both the Gastec tubes and the calorimeter. The calorific value of biogas was also calculated from the formula :

xlorific value (MJ m ')

37.71 (2) 1 = [ 100 -- (CO,('%) + H20('X,)) 100

The calorific value of pure methane is 37.71 hlJ m '. T h e temperature was measured at nine locations within

:he digester with copper,/constantan thermocouples, and the -esults were recorded on data loggers a t 15 min intervals. The pH of the slurry and the digester effluent was deter- nined each week. T h e total solids and volatile solids were

analysed each week (Anon. 1986). T h e volatile fatty acids were quantified by gas chromatography using a Pye Unicam 304 gas chromatograph connected to a Pye Unicam 4810 computer intergrator. T h e column was a 10 m Carbo- wax 20M capillary column (Hewlett Packard) with a film thickness of 1.33 p m and an internal diameter of 0.53 pm. T h e carrier gas, helium, flowed through the column at 40 ml min ~ ' arid the oven temperature was 90°C.

Digester feed

Slurry was hatch fed to the digester daily by gravity flow at 8 m.' d I from a reception pit adjacent to the slurry tank of a slatted beef fattening house for 200 cattle. T h e cattle were fed a diet of concentrates and silage. T h e slurry was mixed with other tarn1 waste and the total solids content of the digester feed was S-lO'M). In order to maintain the digester performance the cattle slurry was often supplemented with hen, pig and potato waste from other units within the farm and from other local farms.

Bacterial enumeration

Escherichra t d f was selected from the digester samples on MacConkey agar (Oxoid CM7). Salmonella typhimurium and 1'. rntrrtrmirtrcu were selected on Desoxycholate Citrate Lactose Sucrose agar (Oxoid CM393) and Yersinia Selec- tive agar (Oxoid C:M653), respectively. Yersinia selective agar containrd an antibiotic mixture (Oxoid SR109) of cef- sulodin, irgasan and novobiocin. Both E. coli and Salm. typhimurium selective agar plates were incubated at 37°C for 24 h whilst the yersinia selective plates were incubated at 30°C for ;I similar period. Campyloboctrr jejuni was selected on Campvlohacter Selective agar (Oxoid CM689) which contained an antibiotic mixture (Oxoid SR117) of poly- myxin R, rrfampicin, trimethoprim lactate and actidone. These plates were incubated under micro-aerophilic conditions at 42 '(1 tor 48 h. Listeria monocytogenes was selected from the digester samples with McBrides's modified List- eria agar ( I m e t t 1988). 'These plates were incubated at 37°C for 24 h.

Background flora grew on the selective plates. Bacterial colonies of E. A'& S d m . typhimurium, Y . rntrrocofitica, L. monocytogenrs and Camp. jqunr were routinely sub-cultured and characterized by the Gram stain reaction, bacterial cell morphology, sodium hippurate hydrolysis and the API 20E

tem. Similarly, the colonies of contami- nating bacteria were routinely sub-cultured and characterized ( K e a r n q r t d. 1993).

Experimental procedure

Four stages tit' the anaerobic digestion process were sampled meeLI1 over a 140 d (20 week) period: (1) slurry at the inlet porr, ( 2 ) digester effluent at the outlet port; (3)

P A T H O G E N S U R V I V A L IN A FARM DIGESTER 217

Table 1 Mean decimal reduction times ( T,,) of pathogenic bacteria during full-scale mesophilic anaerobic digestion

value for each pathogenic bacterium was calculated for the experimental period of 140 d and is shown in Table 1.

~~~ ~ ~

Mean T,, Organism n (4 S.D

Escherichia coli 10 76,9 332.7 SalmoneNa typhimurzum 13 34.5 41.6 Yersrnia enterocolittcu 11 18.2 13.8 Ltsterra monocytogenes 13 28.5 30.3 Campylobacter j qun t 12 4386 342.5

~~~~~~ ~

n, Number of weeks and the number of samples analysed.

separated liquid at the separator and (4) separated solids a t the separator. The separated solids (5.0 g) were suspended in Ringer’s solution (20.0 mi) and mixed. One ml of this suspension was added to 9.0 ml of diluent, mixed and serially diluted to the required dilution. Similarly, 1.0 ml samples of slurry, digested effluent and separated liquid were mixed and serially diluted. Each sample was plated on the respective selective agar plates for E. coli, Salm. typhimurium, Y . enterocolitica, L. monocytogenes and Camp. jejuni and incubated.

Statlstical analysis

The undigested slurry fed daily to the full-scale anaerobic digester continually re-inoculated the digester with pathogenic bacteria. Not surprisingly, the viable numbers of E. coli, Salm. typhimurium, Y . enterocolitica, L. monocytogenes and Camp. jejuni did not decline significantly during full- scale anaerobic digestion. In this study, the specific decline rate (Kd) which is the log,, decline in viable bacteria per day was calculated, from which T,, was subsequently found. In calculating Kd it was assumed that the anaerobic digester was ( 1 ) a continuous system, (2) continually mixed and (3) in a steady state. The rate of pathogen decline (KN) was determined by:

ONou1 = DNi, - K N (3) DN,,,, is the rate of removal of pathogens from the digester effluent and DNi, is the rate of input of pathogens in the slurry. D is the digester dilution rate. I t was assumed that the rate of decline (KN) is proportional to the number of viable cells (N) present (Pirt 1975).

(4)

where 0 is the mean digester retention time.

Kd is expressed as log,. Kd was transformed to T90 by finding its antilog and expressing it as log,,. The time taken for a 90% reduction (T9,) of pathogen numbers was found by dividing the log,, value into 1. The mean T,,

RESULTS

The viable numbers of E. coli, Salm. typhimurium, Y. enterocolitica, L. monocytogenes and Camp. jejuni fluctuated both within and between each stage of the digestion process during the 140 d experimental period. Thus, there are large standard deviations for the mean viable numbers of each bacterium at each stage of the digestion process. Escherichia coli had smaller mean viable counts than any of the other bacteria enumerated at each stage of the digestion process (Table 2); its mean T,, value was 76.9 d (Table 1). After the separation of the solid and liquid fractions of the effluent there was a further reduction in the viable numbers of E. coli. Yersinia enterocolitica was the least resistant bacterium to the anaerobic digester environment during the 140 d experimental period. The viable numbers of Y. enterocolitica declined rapidly during anaerobic digestion ; the mean T90 value was 18.2 d (Table 1) . On average, 4% of the viable population present in slurry remained viable in the digester effluent, but this was reduced further after separating the solid and liquid fractions (Table 2).

Salmonella typhimurium is frequently isolated from cattle and poultry waste, which were fed daily to the full-scale anaerobic digester. This may account for it being the second most viable pathogen enumerated from the slurry input. The viable numbers of Salm. typhimurium in slurry were reduced on average by 82% by anaerobic digestion during the 140 d experimental period. Salmonella typhimurium had a mean T,, value of 34.5 d (Table 1). After the digester effluent was separated, the viable numbers of Salm. typhimurium declined on average by a further 10% in the liquid fraction, but increased in the solid fraction.

In this study L. monocytogenes had the largest viable pathogen population enumerated in the slurry input (Table 2). Silage waste was not fed directly into the digesters during the 140 d experimental period, although silage was fed to the cattle. Listeria monocytogenes had a mean T,, value of 28.5 d.

Campylobacter jejuni was the fourth most numerous pathogen enumerated in the slurry (Table 2). During full scale anaerobic digestion this bacterium had a mean T9, value of 438.6 d. The viable numbers declined on average by 13% during the 140 d experimental period. After separation of the solid fraction from the effluent there was a further reduction in the viable numbers of Camp. jejuni although the viable numbers increased in the liquid fraction during storage (Table 2).

Regression analysis (Sokal and Rohlf 1981) showed that there were no direct correlations between the slurry source, hydraulic retention time (HRT), total solids or the total volatile fatty acid (VFA) concentrations and the decline in

218 THERESA E . KEARNEY ET A L ___. -

Table 2 Arithmetic mean viable counts of the pathogenic bacteria during the full-scale anaerobic digestion process

Arithmetic mean (cfu rnl ’)

Organism Separated Separated

n Slurry Eftluen t liquid solids

Escherirhra coli S.1). (+ )

Sulmnnella i,yphimurrum 13 3.63 x 105 6.72 x 104 3.75 x lo4 1.76 x 105 S.D. (+) x.17 x 105 2.04 x 10’ 4-67 x lo4 5.74 x 105

Yersiniu rn[rrocolitrca 12 2 . 2 ~ 105 9.30 x 10” 5.40 x 103 5.40 x lo3 5.11. (+ ) 6.40 x 1 0 5 9.90 x 10” 6.04 x 103 6.20 x lo3

I,isteriu monncyfogenez 13 2.32 x 106 3.08 x 1 0 5 2.95 x 1 0 5 1 . ~ 2 x 105 S.1). ( + ) 2-65 x 106 3.90 x lo5 3.87 x 105 2.93 x 1 0 5

(.‘amp~~~lubucter jrjunr 12 I 4 4 x 105 9.10 x io4 1.50 x 2.07 x 10“ S.D. ( 4 ) 6.20 x 104 7.30 x 104 7.00 104 1.16 x in4

n , Number of weeks sampled.

viable numbers of I?. coli, Salm. typhimurium, Y . entero- vlzlicu, L. monocytogenes or Camp. jq’uni during the experi- Oncntal period of the full-scale digester. The performance of he full-scale anaerobic digester during the 140 d experi-

,ncntal period is summarized in Table 3.

DISCUSSION

‘The principal cnvironmental factors that influence anaero-

Table 3 Performance of the full-scale anaerobic digester

Parameter n Mean*

llydraulic retention time (d) 20 24.0 2.2 Gas production (m3 d ’) 20 116.6 * 41.3 Methane (X,) I 0 4 x 4 +- 4.0 Calorific value (MJ m - 3 ) 19 18.0 & 1.8 Temperature (“C) 20 2 7 4 * 1.3 O H

slurry eMuent

Loadings (tonnes week- ‘ ) Total solids

slurry effluent

slurry effluent

rota1 VFAs slurry effluent

Volatile solids

19 20

2 0 20

20 20

17 16

7.6 + 0.6 7.7 0.2

* Arithmetic mean S.D. of the data during the 140 d (20 week) s.1 nple period. n Number of weeks sampled and the number of samples mea- sirred; V F A , volatile fatty acid.

bic digester performance are also considered to be important for thc survival of pathogenic bacteria. The daily production of pas fluctuated during this study due to changes in the I I R T and the source of feed to the digester. In laboratory studies Tappouni (1984) found that the maximum hiogas production during semi-continuous digestion at a 7..5 d HRT, corresponded to an increased effect in reducing the numbers of Sulmonella spp. This decline in viable numbers was correlated with increased VFA concentrations and a decrease in pH. Similarly, Kearney (1901) reportcd that Sulm. t.yphiniurium had a T,, value of 2.1 d at a 25 J HRT, but after the HRT was reduced to 5 d thc T9, value was 1.2 d. In this study there were no direct correlations between H R T or the total VFA concentrations in the full-scale digester and the decline in viable pathogen numbers. Other factors such as the source and type of slurry, pH and temperature of the anaerobic digestion process would also have had a considerable effect on the pathogen populations.

In this study the indigenous pathogen population within the full-scak digester had considerably higher mean T,, values than 1 he laboratory attenuated strains within lahor- atory model digesters. In an earlier study (Kearney et al. 1993) the mean T,,, values for a single inoculum of E. coli, Salm. typhimunum, Y . enlerocolitica, L. monocytogrnes and Camp. j q u n t were 1.5, 1.1, 2.5, 35.7 and >71 d respectively during semi-continuous anaerobic digestion. Similarly in this study, Y rnlcrncoliticu had the lowest mean Tgo value during full-scalc anacrobic digestion (Table I), although the value is hignihcantly greater than that obtained in the laboratory digesters. Thus, the magnitude of the T,, values of these path*)genic bacteria appears to differ significantly betwren the laboratory and full-scale anaerobic digesters. The T,, valucs of the pathogenic bacteria within the lahw-

P A T H O G E N S U R V I V A L IN A F A R M DIGESTER 219

atory digesters were based on the decline of viable bacteria in a single inoculum (Kearney et al. 1993). Carrington et al. (1982) reported that the decline of S a l m . duesseldorf fol- lowed first-order kinetics in which the decimal decay rate was 1.6 d - ' . In practice, full-scale anaerobic digesters are run either as semi-continuous or continuous systems in order to be economically viable. Full-scale anaerobic digesters would be re-inoculated each day by slurry due to pathogenic bacteria present. T h e cattle waste fed to the full-scale anaerobic digester was supplemented with hen, pig and potato waste to maintain optimal digester performance. T h e effect this practice had on the overall bacterial flora of the digester input and output is unknown since the bacterial pathogens were not enumerated in the different wastes fed to the digester. Soluble carbohydrates present in slurry are rapidly utilized by the non- methanogenic bacteria and the concentration of such nutri- ents within anaerobic digesters is important in determining the bacterial populations within the system.

T h e conditions in which bacteria grow are important in determining their genotypic and phenotypic characteristics and, hence, their subsequent viability and survival within an environment. I n this study, the decline in viable numbers of E. coli, Salm. typhimurium, L. monocytogenes and C a m p . je juni ranged from 0 to 7 log,, units in 140 d during full-scale digestion. In contrast, these pathogenic bacteria, with the exception of C a m p . je juni , declined by approximately 4 loglo units in 10 d within laboratory model digesters. Genotypic and phenotypic differences may con- tribute to the indigenous pathogen population having significantly higher T,, values than the laboratory strains. In a study by Munch and Schlundt (1983) laboratory strains of E . coli had T,, values of 1-2 d, whilst indigenous faecal E. co/i had significantly higher T,, values of 2.5-5.5 d during continuous rnesophilic digestion. Olsen and Larsen (1987) reported a similar trend in which the indigenous strains of E. col i had higher T,, values than the laboratory strains.

After the separation of the solid and liquid fractions of the effluent there was a further small but variable reduction of the pathogen populations. Jones (1980) proposed that Salmonel la spp. would decline rapidly in the liquid fraction of slurry after being mechanically separated from the solids. In this study the mean viable numbers of S a l m . typhimu- r ium were greater in the separated solid fraction than in the liquid fraction. In contrast, the mean viable numbers of L. monocytogenes and C a m p . j e jun i were greater in the liquid than in the solid fractions.

Reductions of bacterial numbers in farm waste during full-scale anaerobic digestion fluctuated although the indigenous pathogen population survived significantly better than the laboratory strains. T h e decline in viable bacteria during mesophilic anaerobic digestion varied depending on the pathogen species.

ACKNOWLEDGEMENTS

T h e authors wish to thank the Department of Education for Northern Ireland and Farm Gas i t d . (Bishop's Castle, Shropshire, UK) for their financial support during this study.

REFERENCES

Anon. (1986) The Analysis of Agriculture Materials. Ministry of Agriculture, Fisheries and Food Reference book number 427. London : HMSO.

Carrington, E.G., Harman, S.A. and Pike, E.B. (1982) Inac- tivation of Salmonella during anaerobic digestion of sewage sludge. Journal of Applied Bacteriology 53, 33 1-334.

Gadre, R.V., Ranade, D.R. and Godbole, S.H. (1986) A note on survival of salmonellas during anaerobic digestion of cattle dung. Journal of Applied Bacteriology 60, 93-96.

Jones, P.W. (1980) Health hazards associated with the handling of animal wastes. The Veterinary Record 106, 4-7.

Jones, P.W. and Matthews, P.R. (1975) Examination of slurry from cattle for pathogenic bacteria. Journal of Hygiene Cam- bridge 74, 57-64.

Kearney, T.E. ( 1991) Survival of pathogenic bacteria in anaerobic digesters. PhD Thesis. The Queen's University of Belfast, Belfast, UK.

Kearney, T.E., Larkin, M.J. and Levett, P.N. (1993) The effect of slurry storage and anaerobic digestion on survival of pathogenic bacteria. Journal of Applied Bacteriology 74, 8693.

Larsen, H.E. and Munch, B. (1982) Occurrence and survival of pathogenic bacteria in cattle and pig slurry. In Communicable Diseases Resulting from Storage, Handling, Transport and Land- spreading of Manures ed. Watton, J.R. and White, E.G. pp. 161-174. Brussels : Commission of the European Communities.

Lovett, J. (1988) Isolation and enumeration of Listeria monocytogenes. Food Technology 42, 172-175.

Munch, B. and Schlundt, J. (1983) On the reduction of pathogenic and indicator bacteria in animal slurry and sewage sludge subjected to anaerobic digestion or chemical disinfection. In Hygienic Problems of Animal Manures ed. Strauch, D. pp. 131- 149. Stuttgart : University of Hohenheim.

Olsen, J.E. and Larsen, H.E. (1987) Bacterial decimation times in anaerobic digestions of animal slurries. Biological Wastes 21,

Pirt, J.S. (1975) Death of cells in growing cultures. In Principles of Microbe and Cell Cultivation pp. 57-62. Oxford: Blackwell Scientific Publications.

Schlundt, J. (1984) Survival of pathogenic enteric bacteria in anaerobic digestion on slurry-treated land. Dissertation Abstracts International C45 (4). 1025.

Skirrow, M.B. (1990) Foodborne illness: Campylobacter. Lancet

Sokal, R.R. and Rohlf, F.J. (1981) Linear regression. In Biometry, 2nd edn. pp. 499-509. San Francisco: W.H. Freeman.

Tappouni, Y.A. (1984) The fate of Salmonella in anaerobic digestion. PhD Thesis. University College, Cardiff, UK.

Turner, J., Stafford, D.A., Hughes, D.E. and Clarkson, J . (1983) The reduction of three plant pathogens (Fusarium, Corynebac- terium and Globodera) in anaerobic digesters. Agricultural Wastes 6, 1-1 1.

153-168.

(London) 336,921-923.

survival of pathogenic bacteria during mesophilic anaerobic digestion of animal waste

Documents