Declines of zoonotic agents in liquid livestock wastes stored in batches on-farm


Sheryl Avery, Microbiology Research, Direct Laboratory Services Ltd, Wergs Road, Wolverhampton WV6 8TQ, UK


Aim:  To measure the decline rates of zoonotic agents introduced into liquid livestock wastes in on-farm storage tanks.

Methods and Results: Salmonella spp., Escherichia coli O157, Campylobacter jejuni, Listeria monocytogenes and Cryptosporidium parvum, propagated in laboratory-controlled conditions, were inoculated into 35 000-l volumes of fresh livestock wastes (pig slurries, cattle slurries and dirty waters). D-values for bacteria were six to 44 days, and for C. parvum were 133 to 345 days. Campylobacter jejuni declined significantly more rapidly than the other bacterial pathogens, while E. coli O157 declined significantly more slowly. On average, bacterial declines were not affected by the season of waste deposition and storage or by the dry matter content of the wastes, but were more rapid in dirty waters than in pig slurries. The physiciochemical composition of wastes in each category varied significantly.

Conclusions:  Zoonotic agents can survive for several months during storage of liquid livestock wastes. Livestock wastes should be batch-stored and not subjected to continuous additions.

Significance and Impact of the Study:  This study indicates that batches of liquid livestock waste, if contaminated with bacterial pathogens, should be stored for 6 months to reduce contamination levels. Alternative strategies for reducing C. parvum levels in liquid livestock wastes should be explored.


The bacterial pathogens Salmonella, Campylobacter jejuni, Escherichia coli O157, Listeria monocytogenes and the protozoan pathogen Cryptosporidium parvum, occur in a variety of sources, including the gastrointestinal tracts of livestock. Potential routes of pathogen spread on-farm via the contents of livestock gastrointestinal tracts include direct transfer to the gastrointestinal tracts of humans, other livestock, wild animals or insects, and contamination of handlers, equipment, foods or feedstuffs, crops, pasture, waters or air.

One practice on-farm which could facilitate the spread of pathogens from farms into the food chain is spreading pathogen-contaminated livestock wastes onto land that is subsequently used for crop production or livestock grazing (Pell 1997; Jones 1999). A study by Hutchison et al. (2004a) showed that livestock wastes on farms were significantly contaminated with Salmonella, Campylobacter, E. coli O157 or L. monocytogenes at levels up to 2·6 × 108 CFU g−1 or C. parvum at levels up to 3·6 × 103 oocysts g−1. Around 30% of UK livestock wastes harboured at least one of these bacterial pathogens (Hutchison et al. 2004a). Precautions aimed at preventing the spread of livestock-associated pathogens on-farm may be helpful in limiting the spread of livestock-associated pathogens further along the food chain.

Storage is regarded as an effective strategy to reduce pathogen numbers in livestock wastes before being spread on land (Strauch and Ballarini 1994; McGee et al. 2001; Bicudo and Goyal 2003). Storage of solid livestock wastes (farmyard manures) results in a composting effect, with heating from c. 50 to 70°C, which contributes to pathogen reductions in the wastes (Mawdsley et al. 1995; Pell 1997). However, much of the livestock waste from modern intensive farms is liquid and cannot easily be composted (Strauch and Ballarini 1994). Liquid livestock wastes can be categorised as: (i) slurry, which is normally derived from housing or (ii) dirty water, normally derived from milking parlours. Most, but not all, liquid slurry or dirty water wastes are stored on UK farms before being spread on land (Mawdsley et al. 1995; Smith et al. 2000, 2001). Although bacterial numbers decline during storage, bacteria including E. coli, Salmonella and Campylobacter coli survived in slurries for substantial times, ranging up to 20 weeks (Burrows and Rankin 1970; Jones 1976; Kearney et al. 1993; Ajariyakhajorn et al. 1997; Kudva et al. 1998; McGee et al. 2001; Paluszak and Olszewska 2002). The survival of Salmonella (Himathongkham et al. 1999a,b) or E. coli O157 (Wang et al. 1996) during storage of fresh faeces, laboratory-prepared wastes or solid wastes has also been studied. However, how such fresh faeces, laboratory-prepared wastes or solid wastes relate to liquid livestock wastes, which can contain unspecified amounts of faeces, urine, bedding, water, feedstuffs, blood, detergents, antibiotics and nose, throat, mammary gland and vaginal secretions, is not clear (Mawdsley et al. 1995; Pell 1997).

To date, production of microbiologically-based guidelines describing liquid livestock waste management practices to minimize risks of pathogen spread has been confounded by the paucity of data on pathogen declines in liquid livestock wastes in on-farm studies (Pell 1997). There are few data on the fate of pathogens other than Salmonella and E. coli O157 in a variety of liquid wastes currently produced on-farm from different livestock species/types. Also, few studies have exposed pathogens in farm-generated wastes to farm environments with environmental fluctuations. Therefore, the aims of the current study were to investigate declines of five inoculated micro-organisms (Salmonella spp., nonverotoxigenic E. coli O157, L. monocytogenes, Camp. jejuni and C. parvum) during storage in seven types of liquid livestock wastes (slurries or dirty waters) derived from pigs or cattle. The wastes were studied under typical conditions found on-farm (35 000-l volumes, stored in 50 000-l tanks) and were exposed to the on-farm climatic conditions and precipitation that occurred during the study.


Micro-organisms and culture conditions

The bacterial isolates used in the current study (Salmonella Typhimurium, Salmonella Enteritidis, nonverotoxigenic E. coli O157, L. monocytogenes and Camp. jejuni) and the culture conditions used to propagate them in the laboratory have been described previously (Hutchison et al. 2004b). The bacterial pathogens were recent livestock isolates.

The C. parvum strain originated from a cervine animal and had been previously propagated in an ovine host at Moredun Institute (Edinburgh, UK). Oocysts were used directly from the viable cyst suspension provided by Moredun Institute, or when necessary, oocyst numbers were increased by propagation in 14-day old dairy calves. Infection was by feeding milk containing a total dose of 5 × 106 oocysts per day, for three consecutive days. Calf faeces (8 l total volume) were collected for 5 days postinfection, and held in bulk at 4°C for <24 h, until use. Bulk oocyst-containing calf faeces were mixed immediately before use, and contained up to 108 oocysts ml−1 (numbers determined as described below).

Waste collection, inoculation and storage

All wastes were fresh (<72 h postdeposition) and were obtained using waste tankers from animal housing or storage tanks from local farms, during summer or winter. The seven wastes used were: (i) breeder pig slurry; (ii) finishing pig slurry; (iii) dairy cattle slurry with high dry matter (DM); (iv) dairy cattle slurry with low DM; (v) beef cattle slurry with high DM; (iv) dairy cattle slurry with low DM; and (vii) dirty water, largely washed from a dairy parlour. Wastes were stored in 35 000-l batches in 50 m3 above-ground circular tanks, 2·6 m in height and 5 m in diameter. The tanks consisted of a corrugated iron scaffold lined with 4 mm thickness butyl-polythene. Whilst there are no ‘standardized’ slurry tank designs, the tanks used in the current study were typical of those commonly used on UK farms, in both dimensions and construction. Pathogen declines in the stored wastes were measured in studies commencing during both summer and winter, over a 3-year period starting in 2000.

Each of the bacterial pathogens was inoculated into the wastes at c. 106 CFU g−1. That level was chosen to enable the fate of pathogens in the wastes to be studied, and was representative of pathogen levels found in naturally contaminated farm wastes (Hutchison et al. 2004a). Cultured bacteria were introduced directly into each tank and the contents were mixed by vacuum-pumping the inoculated wastes between each tank and waste tanker.

Cryptosporidium parvum oocysts were not released directly into the tanks because it was not possible to propagate the numbers required. Instead, suitable amounts of mixed oocyst suspension or calf faeces, each containing c. 5 × 104 oocysts, were measured into 100 ml polypropylene jars (Ross Labs, Macclesfield, UK). Jars were then filled with waste from each respective tank and the tops were sealed with 0·8 μm membrane (Nucleopore Track-etch, LabSales, Over, UK). Sealed jars were suspended on straw-baling twine into the liquid in the tanks, and the fate of oocysts in the jars was followed. During waste storage, the pH, % DM (w/v), conductivity and ammonium N (amount of N-ions in the waste derived from ammonium) content of waste in the jars was determined as described previously (Hutchison et al. 2004b) and compared with samples taken from the tanks, to determine if any changes in waste chemistry resulted from selective permeability of the 0·8 μm membrane.

After introduction of the zoonotic agents, a Tiny Talk temperature recorder (Orion Components Ltd, Chichester, UK), set to record every 30 minutes, was tied on straw-baling twine and suspended into the centre of each tank. Tanks were covered with a coarse wire mesh but were fully exposed to atmospheric temperature, sunlight and precipitation.

Tank sampling and sample transit

Tanks were not stirred before sampling, as typical on-farm practices were being mimicked and tanks on-farm are usually stirred only occasionally or never (Smith et al. 2001). Samples were taken in triplicate for microbiological and chemical analyses during storage of the liquid livestock wastes. Sampling times are shown in Table 1, although C. parvum sampling was not conducted for some final times. Each bacterial and chemistry sample was comprised of a minimum of five combined subsamples collected from different depths and areas of the tanks. Triplicate samples for C. parvum analysis were taken by removing three oocyst-containing jars from the lower (30 cm from the tank bottom), centre (c. 0·8 m below the waste surface) and upper (20 cm from the waste surface) regions of the tanks. Samples were refrigerated at 2°C and shipped from farm sites to laboratories where analyses commenced within 12 h of sampling. Replicate subsamples were taken on some days to establish the variation in microbiological analyses related to subsampling.

Table 1.  Sampling intervals for microbiological and chemical analyses of liquid livestock wastes during storage on-farm
Livestock waste type and season of deposition and storageSampling time (days of storage)
  1. DM, dry matter content of the waste expressed as a percentage of fresh wet weight.

 Breeder pig slurry0, 2, 4, 8, 16, 32, 64, 94, 108
 Finishing pig slurry0, 2, 4, 8, 16, 32, 64, 94, 108
 Dairy cattle (high DM) slurry0, 1, 2, 4, 8, 16, 32, 112, 182
 Dairy cattle (low DM) slurry0, 1, 2, 4, 8, 16, 32, 112, 182
 Beef cattle (high DM) slurry0, 2, 4, 8, 16, 32, 64, 94, 108
 Beef cattle (low DM) slurry0, 2, 4, 8, 16, 32, 64, 94, 108
 Dirty water0, 1, 2, 3, 8, 16, 32, 112, 182
 Breeder pig slurry0, 2, 4, 8, 16, 31, 64, 93, 156
 Finishing pig slurry0, 2, 4, 8, 16, 31, 64, 93, 156
 Dairy cattle (high DM) slurry0, 2, 4, 8, 16, 32, 65, 93, 165
 Dairy cattle (low DM) slurry0, 2, 4, 8, 16, 32, 65, 93, 165
 Beef cattle (high DM) slurry0, 2, 4, 8, 16, 31, 64, 93, 156
 Beef cattle (low DM) slurry0, 2, 4, 8, 16, 31, 64, 93, 156
 Dirty water0, 2, 4, 8, 16, 32, 65, 93, 165

Microbiological analyses

The bacteriological analyses methods used have been described previously (Hutchison et al. 2004b). Briefly, suspended solids were removed by coarse filtration, and the liquid fraction of the wastes was diluted. Bacteria were captured from the diluted wastes by a second filtration step, and filters were incubated on selective agars. As described previously, Salmonella Typhimurium and Salmonella Enteritidis were not speciated, but were enumerated as genus Salmonella; similarly, Listeria were not speciated, (Hutchison et al. 2004b). Theoretical limits of detection (calculated from dilutions prepared and volumes plated) were 10 CFU ml−1 for most wastes, but were 1 CFU ml−1 for some wastes that were easily filtered. Bacterial counts were calculated from more than one dilution, according to established practice (ISO 4833 1991).

Antibody capture was used to determine the numbers of C. parvum in each jar. Well-mixed waste (1 g) from each jar was vortexed with 10 ml of water-containing 0·01% (v/v) Tween 20 and overlaid onto 40 ml of 1·09 g ml−1 sucrose solution. Each overlaid waste was centrifuged at 5000 g for 10 min without braking. The top 25 ml of each supernatant was mixed with 25 ml of deionized water and recentrifuged as before. Cryptosporidium parvum oocysts were enumerated from each pellet using the GC-combo IMS kit (Dynal Biotech, Wirral, UK) according to manufacturers instructions. Assessment of viability was by 4′,6′-diamidino-2-phenylindole (DAPI) and propidium iodide staining and epifluorescence microscopy as described previously (Pepperell et al. 2003; Hutchison et al. 2004a).

Chemical analyses

The pH, % DM (w/v), conductivity and ammonium N (amount of N-ions in the waste derived from ammonium) content of each waste sample was determined as described previously (Hutchison et al. 2004b).

Analysis of results

Mean pathogen levels and associated standard deviations were normalized by transforming to log10 CFU g−1 using Excel (Excel 2000, Microsoft Corporation, Redmond, WA, USA). D-values (days for a 1-log cycle decrease to occur) were calculated from the first month (31 or 32 days) of bacterial decline curves, and from complete decline curves for C. parvum. D-values were grouped according to organism, waste type, and season of deposition and storage, and compared using t-tests (Excel 2000).


The numbers of all pathogens studied decreased in the wastes during storage, and initial increases in levels were not observed. D-values were calculated from the initial declines (Table 2). Overall, mean Campylobacter declines in the wastes were significantly faster (P < 0·05; D-value 10·7 days) than those of the other bacterial pathogens, while mean E. coli O157 declines were significantly slower (P < 0·05; D-value 21·4 days).

Table 2. D-values (days) for the initially linear declines of zoonotic agents in liquid livestock wastes
Livestock waste type and season of deposition and storageSalmonellaE. coli O157ListeriaCampylobacterC. parvum
  1. DM, dry matter content of the waste expressed as a percentage of fresh wet weight.

  2. –, Indicates that the R2-value of the regression equation was too low (<0·7) to provide a reliable index of initial decline.

 Breeder pig slurry23·520·013·414·4133
 Finishing pig slurry25·625·222·711·4345
 Dairy cattle (high DM) slurry6·86·59·57·1333
 Dairy cattle (low DM) slurry8·911·610·610·4208
 Beef cattle (high DM) slurry18·544·110·915·5250
 Beef cattle (low DM) slurry17·422·413·76·9154
 Dirty water9·08·29·9312
 Mean summer decline rates15·719·713·510·8248
 Breeder pig slurry15·219·116·17·7217
 Finishing pig slurry14·918·914·17·2270
 Dairy cattle (high DM) slurry11·833·39·48·2250
 Dairy cattle (low DM) slurry7·325·416·012·3
 Beef cattle (high DM) slurry16·934·610·815·589
 Beef cattle (low DM) slurry17·018·113·513·0227
 Dirty water11·612·820·010·0
 Mean winter decline rates13·523·214·310·6211
Overall mean decline rates for all wastes and conditions14·621·413·910·7232

The D-values for bacterial pathogens in the cattle slurries ranged from 6·5 to 44·1 days, but were not significantly affected by the DM content of the cattle slurries used (Table 2). Also, the length of time Salmonella survived (up to 93 days) was not affected by the DM content of the cattle slurries. On average, the bacterial pathogens declined more quickly in dirty water (average D-value 11·6 days) than in the pig slurries (D-values of 16·1 and 17·5 days for breeder and finishing pig slurry respectively). Similarly, average declines of the bacterial pathogens in dirty water were more rapid than in beef slurry with a high DM content (average D-values of 11·6 days and 20·9 days respectively).

Cryptosporidium parvum oocysts were stored in membrane-covered jars, but no significant differences were observed between the chemistry of each waste tank and oocyst-containing jars within it (P > 0·05; results not shown). Cryptosporidium parvum survived for extended periods of time in the livestock wastes studied, and calculated rates of decline were much slower than those of the bacterial pathogens (Table 2). For C. parvum, the fastest D-value calculated was 3 months and the slowest almost 1 year (Table 2). In addition, large proportions of the oocysts remained viable throughout the study; a typical C. parvum decline curve is shown (Fig. 1).

Figure 1.

Fate of Cryptosporidium parvum in slurry generated by breeder pigs during summer and stored in 35 000-l batches. bsl00000, total oocysts; bsl00001, viable oocysts; bsl00072, percentage of viable oocysts. Oocysts (5 × 104) were added to polypropylene jars, which were filled with c. 100 ml volumes of slurry. Each jar was sealed at the neck with a semi-permeable membrane (0·8 μm pore size) and sunk into a storage tank containing c. 35 000 l of waste. Data are arithmetic means from triplicate samples; error bars are the associated standard deviations

There was no difference in the mean decline rates of any of the pathogens when stored in summer-deposited wastes compared with winter-deposited wastes. D-values were calculated only from the first 31 or 32 days of each experiment, when average waste temperatures were 12·4 and 4·3°C in studies commencing during summer and winter respectively. The wastes were exposed to ambient temperatures on the farm over periods of several months, so their temperatures did vary with storage time (data not shown). However, the summer- and winter-deposited wastes used had noticeably differing physicochemical properties (Table 3). For example, the ammonium N concentrations in summer- and winter-deposited high DM dairy cattle slurries were 737 and 1623 mg l−1 respectively (Table 3).

Table 3.  Mean physicochemical properties of wastes at the beginning of the experiments
Livestock waste type and season of deposition and storagepHDry matter (% w/v)Conductivity (μSi cm−1)Ammonium N (mg l−1)
  1. DM, dry matter content of the waste expressed as a percentage of fresh wet weight. Data are the means of triplicate samples.

 Breeder pig slurry7·703·2360772930
 Finishing pig slurry7·165·6614333888
 Dairy cattle (high DM) slurry6·907·158976737
 Dairy cattle (low DM) slurry7·172·109903760
 Beef cattle (high DM) slurry6·838·5310731827
 Beef cattle (low DM) slurry7·032·964047480
 Dirty water6·201·136710991
 Breeder pig slurry7·566·00363751
 Finishing pig slurry7·6310·505683777
 Dairy cattle (high DM) slurry7·5710·0087571623
 Dairy cattle (low DM) slurry7·732·208207959
 Beef cattle (high DM) slurry7·506·004581747
 Beef cattle (low DM) slurry7·602·20200993
 Dirty water5·400·904133303

In some of the wastes, bacterial declines were not log-linear over the entire storage times, but significant tailing occurred. For example, low levels (5 CFU ml−1) of viable E. coli O157 remained in the high DM beef cattle slurry when the summer experiment was terminated after 108 days (Fig. 2). Small numbers of viable Salmonella (18 CFU ml−1) also remained in the summer breeder pig slurry after 108 days (Fig. 2).

Figure 2.

Decline of bacterial pathogens in (a) high DM beef cattle slurry generated in summer and (b) breeder pig slurry generated in summer stored in 35 000-l batches. •, Salmonella spp.; ○, Escherichia coli O157; bsl00072, Listeria; bsl00083, Campylobacter. Data points are geometric means of triplicate samples; error bars are the associated standard deviations

The variation in measured bacterial levels between subsamples taken on any 1 day was significant. Replicate analyses of waste subsamples from tanks immediately after inoculation and mixing produced a range of almost three logs between the highest and lowest results, and coefficients of variation of up to 138% (data not shown). In contrast, the physicochemical properties of the wastes did not vary significantly between replicate samples taken on each sampling day (data not shown).


The range of D-values for bacterial pathogens in cattle slurries in the current study (c. 6–44 days) may not be directly comparable with D-values calculated in other studies, for reasons discussed below. However, in cattle slurries stored at 4°C, Camp. coli had a D-value of 17·6 days (Paluszak and Olszewska 2002), while Camp. jejuni had a D-value of >112 days (Kearney et al. 1993). Kearney et al. (1993) also calculated D-values of 17–22 days and 29 to >84 days for Salmonella Typhimurium and L. monocytogenes, respectively, in cattle slurries at temperatures of 4 and 17°C.

Previously, it was reported that Salmonella Dublin survived longer in cattle slurry adjusted to 5% DM than in 1% DM slurry, in laboratory-scale (400 ml) experiments (Jones 1976). In contrast, Salmonella did not survive longer in higher DM rather than in lower DM cattle slurries in the current study. The phenomenon of drying itself (rather than DM content) is probably more influential on bacterial behaviour in solid livestock wastes (farmyard manures) (Himathongkham et al. 1999b). It was noticeable in the current study that dirty waters caused more rapid death of the bacterial pathogens than some of the other wastes. Dirty water is comprised largely of milk parlour washings, and normally contains detergents, sanitizers and possibly veterinary antibiotic residues, any of which could hasten bacterial declines (Mawdsley et al. 1995).

In the current study, C. parvum oocysts were stored in membrane-covered jars, as sufficient concentrations of oocysts could not be generated to inoculate, at suitable levels, the 35 000-l volumes of liquid wastes used. However, the membranes remained permeable to ammonium N and H+ ions throughout the course of the study and the waste in the jars appeared equivalent (with respect to these ions, plus % DM and conductivity) to the waste in the main body of the tanks. The results show that treatment of C. parvum oocyst-contaminated liquid livestock wastes by batch storage alone is not a practical solution for reducing oocyst levels, because very extended storage times are required for significant declines. In contrast, previous laboratory-based research has shown that significantly more rapid reductions in C. parvum oocyst levels occurred when liquid animal wastes were aerated, as the temperature and ammonia content both increased markedly (Kemp et al. 1995). Therefore, strategies such as these may help reduce C. parvum levels in batch-stored liquid wastes. Nevertheless, other studies showed that, in practice, on UK farms, 40% of liquid pig waste stores (Smith et al. 2000) and 10% of liquid cattle waste stores (Smith et al. 2001) were never aerated. In addition, liquid waste storage on farms is a concern for farmers as most do not have the capacity to store more than a few months of waste (Smith et al. 2000, 2001). Usually, therefore, UK slurry tanks and lagoons, are not operated as batch stores and are subject to regular and continuous additions of fresh waste (Hutchison et al. 2000; Smith et al. 2001; Hutchison et al. 2005).

Laboratory studies have shown that Salmonella or E. coli O157 survive better in faeces or livestock waste stored at lower, rather than higher, temperatures (Ajariyakhajorn et al. 1997; Kudva et al. 1998; Himathongkham et al. 1999a; Olson et al. 1999). However, in the current study, it is likely that other factors contributed to the different bacterial declines observed in the wastes deposited and stored in the different seasons, as their physicochemical properties were different. The variability of livestock wastes has been noted previously (Hutchison et al. 2004a), and may depend partly on stock diet (Patni and Jui 1991; Nahm 2003). Aqueous NH3+ concentration (Himathongkham and Riemann 1999) or pH (Ajariyakhajorn et al. 1997; Park and Diez-Gonzalez 2003) could influence bacterial declines in livestock wastes. Because animal wastes differ markedly, it does not necessarily follow that observations based on one batch of waste would apply to all batches of identically-named waste (Jones 1976). In the current study, it was not possible to store livestock wastes from 1 year to the next in a manner which would have prevented physicochemical and microbiological changes and ensured continuity of this complex substrate throughout the study.

One aspect of this study that merits mention is sampling error. The numbers of bacterial pathogens we found in the replicate subsamples taken on some days varied greatly. In contrast, the triplicate waste samples, taken on those same days, were not significantly different with respect to their physicochemical properties. At the time of inoculation, we attempted to mix the bacteria into the slurry tanks thoroughly. However, a balance had to be made between homogeneous distribution of bacteria throughout the tanks, and uncharacteristic aeration of the wastes, which has been shown to cause exceptionally rapid declines of bacteria in livestock wastes (Heinonen-Tanski et al. 1998; Kudva et al. 1998). Also, we avoided stirring before sampling to mimic on-farm practice and to avoid excessive aeration. Therefore, the large variations we observed in pathogen numbers, which were probably partially because of our use of large farm-scale tanks and mimicking of commercial practices, could have masked otherwise significant relationships.

In the current study, bacterial decline rates were not log-linear for some of the pathogen-waste combinations over the course of the study. McGee et al. (2001) also found significant tailing of E. coli O157:H7 in one cattle slurry they examined, but not in the other. Log linear bacterial declines were reported in stored cattle slurry (Kearney et al. 1993; McGee et al. 2001) or diluted faeces (Himathongkham et al. 1999a). Although we do not know the reasons for the observation of tailing observed in the current study, there are some plausible explanations. First, the bacterial pathogens may not have been distributed homogenously throughout the tanks, for reasons discussed above. This would be compounded if bacteria exhibited active motility towards specific areas in the tanks. Secondly, the effect may be an artefact of bacterial behaviour in the wastes and the quantitative bacterial methods used. Bacteria may become aggregated after attaching to each other or to small particles in the waste. Aggregated bacteria would produce single colonies, even after the death of some individual cells in the aggregate, if laboratory methods failed to separate them. If significant numbers of aggregated bacteria were present in the wastes, and being closely associated conferred an advantage so that they survived longer than planktonic cells, then death could have been log-linear but not observed as such. Different waste types, and indeed different batches of the same waste type are likely to cause bacterial aggregation to differing degrees which may explain why the effect was observed only in some wastes. Thirdly, natural heterogeneity of the pure culture inocula and/or of bacteria persisting in the wastes over time could contribute to some individual cells adapting to the waste environment over the course of the experiment, and surviving for longer than other cells. Irrespective of reasons for the tailing, it is an important consideration when determining the length of waste storage times which are required for pathogen depletion. However, the length of time for pathogen depletion in livestock wastes is perceived to be of less importance than the rate of decline, as survival time is partly dependent on initial pathogen concentration (Pell 1997).

In the UK, initial draft guidance to farmers, intended to reduce microbial levels in livestock wastes before they are spread on land to grow ready-to-eat crops, recommended that slurries be anaerobically digested or treated with lime (Anon 2002). The current study has confirmed that passive storage of C. parvum-contaminated liquid livestock waste is not an effective strategy for reducing levels of this protozoan pathogen. The results show that 6 months would be an effective time, including an inbuilt safety margin, for storing liquid livestock wastes on-farm to ensure that any bacterial pathogens present have declined below detectable levels. This interval assumes that during storage, no further additions of fresh liquid waste are made to the store. The current study has contributed scientifically to the updating of draft guidance to farmers, on reducing microbial contamination levels in organic wastes (Anon 2002). The guidance, currently under consultation, now recommends that batches of liquid livestock wastes are stored for at least 6 months, or are treated with lime for at least 3 months, before being spread on land where the intention is to grow ready-to-eat crops (F. Opesan, personal communication).


The authors thank Barbara Rarata and Nikki Mason, who provided technical assistance. Rob Davies, Veterinary Laboratories Agency, UK generously provided recent livestock isolates of Salmonella. The United Kingdom Food Standards Agency (UK-FSA) provided prepublication UK-FSA livestock waste use guidance drafts. Richard Lavern, Scottish Agricultural College, St Boswells propagated oocysts in neonate calves. Barbara Rarata kindly translated publications from Polish. This study was funded by the Organic Wastes Programme B17 of the UK-FSA, project BO5003.