Reduction of Experimental Salmonella and Campylobacter Contamination of Chicken Skin by Application of Lytic Bacteriophages

Lytic bacteriophages, applied to chicken skin that had been experimentally contaminated with Salmonella enterica serovar Enteritidis or Campylobacter jejuni at a multiplicity of infection (MOI) of 1, increased in titer and reduced the pathogen numbers by less than 1 log₁₀ unit. Phages applied at a MOI of 100 to 1,000 rapidly reduced the recoverable bacterial numbers by up to 2 log₁₀ units over 48 h. When the level of Salmonella contamination was low (< log₁₀ 2 per unit area of skin) and the MOI was 10⁵, no organisms were recovered. By increasing the number of phage particles applied (i.e., MOI of 10⁷), it was also possible to eliminate other Salmonella strains that showed high levels of resistance because of restriction but to which the phages were able to attach.     

Survival of Salmonella on refrigerated chicken carcasses and subsequent transfer to cutting board

Objective:  To determine the effect of refrigeration time and temperature on Salmonella cell numbers on inoculated chicken carcasses and their transfer to a plastic cutting board.
Methods and Results:  The survival of Salmonella on chicken skin and the transfer to a plastic cutting board when exposed to different refrigeration temperatures (2, 6 or 8°C) for 9 days were the two main issues on which this work focused. Two scenarios were carried out to ascertain these effects: carcasses treated with a decontaminating acetic acid solution and untreated carcasses. All of the contaminated carcasses remained contaminated after 9 days of refrigeration. However, on untreated samples, while Salmonella numbers increased almost 1·5 log at 8°C, the pathogen numbers decreased about 1 log at 2 and 6°C. On acid-treated samples, cell numbers slightly decreased at all of the temperatures studied. Temperature did not affect salmonellae transfer to the cutting board, but time did. Acid decontamination increased cell numbers transferred to the cutting board compared with untreated samples.
Conclusion:  Proper refrigeration at low temperatures did not allow Salmonella numbers to rise, regardless of which carcasses had been, or had not been, acid treated. Despite the fact that the rate of transfer was not affected by temperature, the acid treatment detached Salmonella cells from the chicken skin and, therefore, the probability of greater cross-contamination should be studied further.
Significance and Impact of the Study:  The results of this study may provide better information about the refrigeration conditions for fresh chicken storage and also determine if these, along with acetic acid decontamination of broiler chicken, would affect the pathogen transfer to a cutting board.     

Washing and chilling as critical control points in pork slaughter hazard analysis and critical control point (HACCP) systems

AIMS: The aim of this research was to examine the effects of preslaughter washing, pre-evisceration washing, final carcass washing and chilling on final carcass quality and to evaluate these operations as possible critical control points (CCPs) within a pork slaughter hazard analysis and critical control point (HACCP) system. METHODS AND RESULTS: This study estimated bacterial numbers (total viable counts) and the incidence of Salmonella at three surface locations (ham, belly and neck) on 60 animals/carcasses processed through a small commercial pork abattoir (80 pigs d(-1)). Significant reductions (P < 0.05) in bacterial numbers were noted at some stages of the slaughter/dressing process, i.e. the process of hair removal (scalding-dehairing and singeing) resulted in an approx. 4.5 log10 cfu cm(-2) decrease in bacterial numbers. A significant increase (P < 0.05) in bacterial numbers was observed after pre-evisceration washing. Final washing increased the bacterial counts to between 3.6 and 3.8 log10 cfu cm(-2) while chilling effected a small but statistically significant (P < 0.05) increase to between 4.5 and 4.7 log10 cfu cm(-2). The incidence of Salmonella on pigs at the farm was 27%, decreasing to 10% after preslaughter washing. However, stunning and bleeding effected a considerable increase in Salmonella contamination and the incidence after these operations was 50%, which was reduced to 0% during the scalding-dehairing process. CONCLUSIONS: Washing the live animals and subsequent carcasses with cold water is not an effective control measure but chilling may be used as a CCP. SIGNIFICANCE AND IMPACT OF THE STUDY: Recent changes in European Union legislation legally mandate HACCP in pork slaughter plants. This research will provide a sound scientific basis on which to develop and implement effective HACCP in pork abattoirs.     

Application of antimicrobial ice for reduction of foodborne pathogens (Escherichia coli O157:H7, Salmonella Typhimurium, Listeria monocytogenes) on the surface of fish

AIMS: The efficacy of antimicrobial ice was evaluated for the reduction of foodborne pathogens on the surface of fish. METHODS AND RESULTS: Antimicrobial ice containing chlorine dioxide (ClO2) was utilized to control foodborne pathogens in laboratory media and on fish skin. Escherichia coli O157:H7, Salmonella serotype Typhimurium and Listeria monocytogenes strains were treated with antimicrobial ice for 30 min on plates of selective agar and for 120 min on fish skin at room temperature, and then incubated for enumeration. After treatment with 100 ppm ClO2 for 30 min, 5.4, 4.4 and 3.2 log10 reduction was obtained with E. coli O157:H7, Salm. Typhimurium and L. monocytogenes on laboratory media, respectively. When antimicrobial ice (100 ppm ClO2) was applied to fish skin for 120 min, total reduction of E. coli O157:H7, Salm. Typhimurium and L. monocytogenes was 4.8, 2.6 and 3.3 log10, respectively. CONCLUSION: The initial load of foodborne pathogens was reduced by antimicrobial ice and the lowered microbial level was maintained during treatment. SIGNIFICANCE AND IMPACT OF THE STUDY: The application of antimicrobial ice is a simple and effective method for the safe preservation of fish.     

Bacteriophage T4 multiplication in a glucose-limited Escherichia coli biofilm

An Escherichia coli K-12 biofilm was grown at a dilution rate of 0.028 h(-1) for 48 h in a glucose-limited chemostat coupled to a modified Robbins' device to determine its susceptibility to infection by bacteriophage T4. Bacteriophage T4 at a multiplicity of infection (MOI) of 10 caused a log reduction in biofilm density (expressed as colony forming units (CFU) per cm2) at 90 min postinfection. After 6 h, a net decrease and equilibrium in viral titer was seen. When biofilms were exposed to T4 phage at a MOI of 100, viral titer doubled after 90 min. After 6 h, viral titers (expressed as plaque forming units (PFU) per cm2) stabilized at levels approximately one order of magnitude higher than seen at a MOI of 10. Scanning confocal laser microscopy images also indicated disruption of biofilm morphology following T4 infection with the effects being more pronounced at a MOI of 100 than at a MOI of 10. These results imply that biofilms under carbon limitation can act as natural reservoirs for bacteriophage and that bacteriophage can have some influence on biofilm morphology.     

Survival of Salmonella in peanut butter and peanut butter spread

In 1996, the first documented outbreak of salmonellosis associated with the consumption of peanut butter was reported. This study was undertaken to determine survival characteristics of high (5.68 log10 cfu g(-1)) and low (1.51 log10 cfu g(-1)) inocula of a five-serotype mixture of Salmonella in five commercial peanut butters and two commercial peanut butter spreads. Populations in samples inoculated with 5.68 log10 cfu g(-1) and stored for 24 weeks at 21 or 5 degrees C decreased 4.14-4.50 log10 cfu g(-1) and 2.86-4.28 log10 cfu g(-1), respectively, depending on the formulation. The order of retention of viability was: peanut butter spreads > traditional (regular) and reduced sugar, low-sodium peanut butters > natural peanut butter. Differences in rates of inactivation are attributed to variation in product composition as well as size and stability of water droplets in the colloidal matrix, which may influence nutrient availability. With the exception of natural peanut butter, products initially inoculated with 1.51 log10 cfu of Salmonella g(-1) (32 cfu g(-1)) were positive for the pathogen after storage for 24 weeks at 5 degrees C. At 21 degrees C, however, with the exception of one peanut butter spread, all products were negative for Salmonella after storage for 24 weeks. Post-process contamination of peanut butter and spreads with Salmonella may to result in survival in these products for the duration of their shelf life at 5 degrees C and possibly 21 degrees C, depending on the formulation.     

Enhanced inactivation of Listeria monocytogenes by nisin in the presence of ethanol

AIMS: The effect of combinations of nisin and ethanol on the survival of Listeria monocytogenes was investigated. METHODS AND RESULTS: Killing by nisin was enhanced during simultaneous exposure to ethanol (2-7% v/v). For example, while 10 IU ml(-1) nisin reduced viability by 1 log unit in 20 min, a combination of this antimicrobial peptide and 5% ethanol, reduced numbers of surviving cells by 3 log units. Increasing the concentrations of either ethanol (2-7%) or nisin (10-50 IU ml(-1)) led to increased cell death with synergy being demonstrated for all combinations tested and at a range of temperatures from 5 to 37 degrees C. CONCLUSIONS: Ethanol can act synergistically with nisin to reduce the survival of L. monocytogenes. SIGNIFICANCE AND IMPACT OF THE STUDY: Combinations of ethanol and nisin may be feasible as an effective way of controlling this pathogen in the food processing environment.     

Comparative survival of free shiga toxin 2-encoding phages and Escherichia coli strains outside the gut

The behavior outside the gut of seeded Escherichia coli O157:H7, naturally occurring E. coli, somatic coliphages, bacteriophages infecting O157:H7, and Shiga toxin 2 (Stx2)-encoding bacteriophages was studied to determine whether the last persist in the environment more successfully than their host bacteria. The ratios between the numbers of E. coli and those of the different bacteriophages were clearly lower in river water than in sewage of the area, whereas the ratios between the numbers of the different phages were similar. In addition, the numbers of bacteria decreased between 2 and 3 log units in in situ survival experiments performed in river water, whereas the numbers of phages decreased between 1 and 2 log units. Chlorination and pasteurization treatments that reduced by approximately 4 log units the numbers of bacteria reduced by less than 1 log unit the numbers of bacteriophages. Thus, it can be concluded that Stx2-encoding phages persist longer than their host bacteria in the water environment and are more resistant than their host bacteria to chlorination and heat treatment.     

Evaluation of lactic acid bacterium fermentation products and food-grade chemicals to control Listeria monocytogenes in blue crab (Callinectes sapidus) meat.

Fresh blue crab (Callinectes sapidus) meat was obtained from retail markets in Florida and sampled for viable Listeria monocytogenes. The pathogen was found in crabmeat in three of four different lots tested by enrichment and at levels of 75 CFU/g in one of the same four lots by direct plating. Next, crabmeat was steam sterilized, inoculated with a three-strain mixture of L. monocytogenes (ca. 5.5 log10 CFU/g), washed with various lactic acid bacterium fermentation products (2,000 to 20,000 arbitrary units [AU]/ml of wash) or food-grade chemicals (0.25 to 4 M), and stored at 4 degrees C. Counts of the pathogen remained relatively constant in control samples during storage for 6 days, whereas in crabmeat washed with Perlac 1911 or MicroGard (10,000 to 20,000 AU), numbers initially decreased (0.5 to 1.0 log10 unit/g) but recovered to original levels within 6 days. Numbers of L. monocytogenes cells decreased 1.5 to 2.7 log10 units/g of crabmeat within 0.04 day when washed with 10,000 to 20,000 AU of Alta 2341, enterocin 1083, or Nisin per ml. Thereafter, counts increased 0.5 to 1.6 log10 units within 6 days. After washing with food-grade chemicals, modest reductions (0.4 to 0.8 log10 unit/g) were observed with sodium acetate (4 M), sodium diacetate (0.5 or 1 M), sodium lactate (1 M), or sodium nitrite (1.5 M). However, Listeria counts in crabmeat washed with 2 M sodium diacetate decreased 2.6 log10 units/g within 6 days. In addition, trisodium phosphate reduced L. monocytogenes counts from 1.7 (0.25 M) to > 4.6 (1 M) log10 units/g within 6 days.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparative Survival of Free Shiga Toxin 2-Encoding Phages and Escherichia coli Strains outside the Gut

The behavior outside the gut of seeded Escherichia coli O157:H7, naturally occurring E. coli, somatic coliphages, bacteriophages infecting O157:H7, and Shiga toxin 2 (Stx2)-encoding bacteriophages was studied to determine whether the last persist in the environment more successfully than their host bacteria. The ratios between the numbers of E. coli and those of the different bacteriophages were clearly lower in river water than in sewage of the area, whereas the ratios between the numbers of the different phages were similar. In addition, the numbers of bacteria decreased between 2 and 3 log units in in situ survival experiments performed in river water, whereas the numbers of phages decreased between 1 and 2 log units. Chlorination and pasteurization treatments that reduced by approximately 4 log units the numbers of bacteria reduced by less than 1 log unit the numbers of bacteriophages. Thus, it can be concluded that Stx2-encoding phages persist longer than their host bacteria in the water environment and are more resistant than their host bacteria to chlorination and heat treatment.     

Enteric bacteria in aerobically digested sludge

Indicator bacteria, Salmonella spp., and total aerobic bacteria were determined in samples of undigested sludge and sludge that had been treated by one or two stages of aerobic digestion. Aerobic sludge digestion reduced the level of indicator bacteria by 1 to 2 log10 per g. The level of Salmonella spp. was also reduced during aerobic treatment of sludge. In general, aerobic treatment of sludge reduced, but did not eliminate, indicator bacteria and Salmonella spp.     

Enteric bacteria in aerobically digested sludge.

Indicator bacteria, Salmonella spp., and total aerobic bacteria were determined in samples of undigested sludge and sludge that had been treated by one or two stages of aerobic digestion. Aerobic sludge digestion reduced the level of indicator bacteria by 1 to 2 log10 per g. The level of Salmonella spp. was also reduced during aerobic treatment of sludge. In general, aerobic treatment of sludge reduced, but did not eliminate, indicator bacteria and Salmonella spp.     

Reduction of Escherichia coli O157:H7 by stimulated Pediococcus acidilactici

This study was conducted to determine if stimulating the growth of meat starter culture (Pediococcus acidilactici) in a laboratory medium (Brain Heart Infusion broth +2.3% NaCl + 1.5% sucrose; LBHI) and during meat fermentation would control Escherichia coli O157:H7. In LBHI medium without P. acidilactici, the numbers of E. coli O157:H7 increased from 4.00 to 8.34 log10 cfu ml-1, whereas in the presence of P. acidilactici (approximately 6.0 log10 cfu ml-1) in LBHI, LBHIM (LBHI + 0.005% MnSO4), LBHIO (LBHI + 0.3 unit ml-1 Oxyrase), and LBHIMO (LBHI + M + O), the numbers of E. coli O157:H7 increased from 4.00 to 8.05, 7.50, 7.99, and 6.50 log10 cfu ml-1, respectively, after incubation at 40 degrees C for 15 h. During salami fermentation, the numbers of E. coli O157:H7 changed from 7.00 to 6.40 and 5.10 log10 cfu g-1 without and with P. acidilactici (approximately 7.0 log10 cfu g-1), respectively. Stimulated P. acidilactici by M, O, and MO further reduced the number of E. coli O157:H7 from 7.00 to 4.00, 4.80, and 3.65 log10 cfu g-1, respectively. The combination of MO was a better growth stimulator for P. acidilactici, which controlled E. coli O157:H7 in both systems (P < 0.05).     

Inactivation of animal viruses during sewage sludge treatment

Using a previously developed filter adsorption technique, the inactivation of a human rotavirus, a coxsackievirus B5, and a bovine parvovirus was monitored during sludge treatment processes. During conventional anaerobic mesophilic digestion at 35 to 36 degrees C, only minor inactivation of all three viruses occurred. The k' values measured were 0.314 log10 unit/day for rotavirus, 0.475 log10 unit/day for coxsackievirus B5, and 0.944 log10 unit/day for parvovirus. However, anaerobic thermophilic digestion at 54 to 56 degrees C led to rapid inactivation of rotavirus (k' greater than 8.5 log10 units/h) and of coxsackievirus B5 (k' greater than 0.93 log10 unit/min). Similarly, aerobic thermophilic fermentation at 60 to 61 degrees C rapidly inactivated rotavirus (k' = 0.75 log10 unit/min) and coxsackievirus B5 (k' greater than 1.67 log10 units/min). Infectivity of parvovirus, however, was only reduced by 0.213 log10 unit/h during anaerobic thermophilic digestion and by 0.353 log10 unit/h during aerobic thermophilic fermentation. Furthermore, pasteurization at 70 degrees C for 30 min inactivated the parvovirus by 0.72 log10 unit/30 min. In all experiments the contribution of temperature to the total inactivation was determined separately and was found to be predominant at process temperatures above 54 degrees C. In conclusion, the most favorable treatment to render sludge hygienically safe from the virological point of view would be a thermal treatment (60 degrees C) to inactivate thermolabile viruses, followed by an anaerobic mesophilic digestion to eliminate thermostable viruses that are more sensitive to chemical and microbial inactivations.     

Inactivation of animal viruses during sewage sludge treatment.

Using a previously developed filter adsorption technique, the inactivation of a human rotavirus, a coxsackievirus B5, and a bovine parvovirus was monitored during sludge treatment processes. During conventional anaerobic mesophilic digestion at 35 to 36 degrees C, only minor inactivation of all three viruses occurred. The k' values measured were 0.314 log10 unit/day for rotavirus, 0.475 log10 unit/day for coxsackievirus B5, and 0.944 log10 unit/day for parvovirus. However, anaerobic thermophilic digestion at 54 to 56 degrees C led to rapid inactivation of rotavirus (k' greater than 8.5 log10 units/h) and of coxsackievirus B5 (k' greater than 0.93 log10 unit/min). Similarly, aerobic thermophilic fermentation at 60 to 61 degrees C rapidly inactivated rotavirus (k' = 0.75 log10 unit/min) and coxsackievirus B5 (k' greater than 1.67 log10 units/min). Infectivity of parvovirus, however, was only reduced by 0.213 log10 unit/h during anaerobic thermophilic digestion and by 0.353 log10 unit/h during aerobic thermophilic fermentation. Furthermore, pasteurization at 70 degrees C for 30 min inactivated the parvovirus by 0.72 log10 unit/30 min. In all experiments the contribution of temperature to the total inactivation was determined separately and was found to be predominant at process temperatures above 54 degrees C. In conclusion, the most favorable treatment to render sludge hygienically safe from the virological point of view would be a thermal treatment (60 degrees C) to inactivate thermolabile viruses, followed by an anaerobic mesophilic digestion to eliminate thermostable viruses that are more sensitive to chemical and microbial inactivations.     

Characterization of bacteriophages with a lytic effect on various Salmonella serotypes and Escherichia coli O157:H7

Four phages isolated from cattle and poultry feces were analyzed for their ability to lyse Salmonella serotypes and Escherichia coli O157:H7. The phage one-step growth curves, morphology, and genetic characteristics were determined. All phages showed a lytic effect on various Salmonella serotypes and E. coli O157:H7, which lysed at least 70% of the 234 strains tested. The phages had latent periods ranging from 10 to 15 min and generation times of 30 to 45 min, while burst size fluctuated between 154 and 426 PFU/cell. Phages morphology showed isometric and elongated heads and rigid contractile tails, consistent with morphology of the Myoviridae family. Phages' DNA dendrograms showed a distinctive RFLP when digested by HindIII and EcoRV, and SDS-PAGE profile showed distinctive proteins expression as well. In vitro phage challenge showed a total reduction of E. coli O157:H7, Salmonella Typhimurium and Saintpaul counts at 2 h, whereas for Salmonella Montevideo a reduction and retardation growth, at a multiplicity of infection (MOI) of 100, was observed; however, under a MOI of 10 000, no viable cells were detected after 4 h. The wide host ranges of these phages suggested they could be used for simultaneous biocontrol of some Salmonella serotypes and E. coli O157:H7.     

Control of Listeria spp. by competitive-exclusion bacteria in floor drains of a poultry processing plant

In previous studies workers determined that two lactic acid bacterium isolates, Lactococcus lactis subsp. lactis C-1-92 and Enterococcus durans 152 (competitive-exclusion bacteria [CE]), which were originally obtained from biofilms in floor drains, are bactericidal to Listeria monocytogenes or inhibit the growth of L. monocytogenes both in vitro and in biofilms at 4 to 37 degrees C. We evaluated the efficacy of these isolates for reducing Listeria spp. contamination of floor drains of a plant in which fresh poultry is processed. Baseline assays revealed that the mean numbers of Listeria sp. cells in floor drains sampled on six different dates (at approximately biweekly intervals) were 7.5 log(10) CFU/100 cm(2) for drain 8, 4.9 log(10) CFU/100 cm(2) for drain 3, 4.4 log(10) CFU/100 cm(2) for drain 2, 4.1 log(10) CFU/100 cm(2) for drain 4, 3.7 log(10) CFU/100 cm(2) for drain 1, and 3.6 log(10) CFU/100 cm(2) for drain 6. The drains were then treated with 10(7) CE/ml in an enzyme-foam-based cleaning agent four times in 1 week and twice a week for the following 3 weeks. In samples collected 1 week after CE treatments were applied Listeria sp. cells were not detectable (samples were negative as determined by selective enrichment culture) for drains 4 and 6 (reductions of 4.1 and 3.6 log(10) CFU/100 cm(2), respectively), and the mean numbers of Listeria sp. cells were 3.7 log(10) CFU/100 cm(2) for drain 8 (a reduction of 3.8 log(10) CFU/100 cm(2)), <1.7 log(10) CFU/100 cm(2) for drain 1 (detectable only by selective enrichment culture; a reduction of 3.3 log(10) CFU/100 cm(2)), and 2.6 log(10) CFU/100 cm(2) for drain 3 (a reduction of 2.3 log(10) CFU/100 cm(2)). However, the aerobic plate counts for samples collected from floor drains before, during, and after CE treatment remained approximately the same. The results indicate that application of the two CE can greatly reduce the number of Listeria sp. cells in floor drains at 3 to 26 degrees C in a facility in which fresh poultry is processed.     

Fate of Cryptosporidium oocysts, Giardia cysts, and microbial indicators during wastewater treatment and anaerobic sludge digestion.

The extent of reduction in selected microorganisms was tested during both aerobic wastewater treatment and anaerobic digestion of sludge at the wastewater treatment plant in Ottawa to compare the removal of two encysted pathogenic protozoa with that of microbial indicators. Samples collected included the raw wastewater, the primary effluent, the treated wastewater, the mixed sludge, the decanted liquor, and the cake. All of the raw sewage samples were positive for Cryptosporidium oocysts and Giardia cysts, as well as for the other microorganisms tested. During aerobic wastewater treatment (excluding the anaerobic sludge digestion), Cryptosporidium and Giardia were reduced by 2.96 log10 and 1.40 log10, respectively. Clostridium perfringens spores, Clostridium perfringens total counts, somatic coliphages, and heterotrophic bacteria were reduced by approximately 0.89 log10, 0.96 log10, 1.58 log10, and 2.02 log10, respectively. All of the other microorganisms were reduced by at least 3.53 log10. Sludge samples from the plant were found to contain variable densities of microorganisms. Variability in microbial concentrations was sometimes great between samples, stressing the importance of collecting a large number of samples over a long period of time. In all cases, the bacterial concentrations in the cake (dewatered biosolids) samples were high even if reductions in numbers were observed with some bacteria. During anaerobic sludge digestion, no statistically significant reduction was observed for Clostridium perfringens, Enterococcus sp., Cryptosporidium oocysts, and Giardia cysts. A 1-2 log10 reduction was observed with fecal coliforms and heterotrophic bacteria. However, the method utilized to detect the protozoan parasites does not differentiate between viable and nonviable organisms. On the other hand, total coliforms and somatic coliphages were reduced by 0.35 log10 and 0.09 log10, respectively. These results demonstrate the relative persistence of the protozoa in sewage sludge during wastewater treatment.     

Impact of dried skim milk in production diets on Lactobacillus and pathogenic bacterial shedding in growing-finishing swine

AIMS: To determine the possible effects of inclusion of dried skim milk (DSM) in swine diets on indigenous Lactobacillus spp. and Escherichia coli, and its potential for controlling pathogen shedding and affect animal growth in growing-finishing swine. METHODS AND RESULTS: Animals were fed over three dietary phases to match production needs from age 10-14 weeks, 14-18 weeks and 18-22 weeks. For each feeding phase, diets were formulated to contain 0 or 10% DSM (balanced for metabolizable energy and true ileal digestible amino acids). Animals were weighed every 2 weeks and faecal samples were collected from 40 animals (20 with DSM and 20 without DSM) at week 10 (d 0 on diets), 14, 18 and 22 of age, and were analysed for Lactobacillus spp., Enterobacteriaceae, coliforms, E. coli, Salmonella, Campylobacter and E. coli O157:H7. At the start of the study (week 10), faecal bacterial counts (log10 CFU g(-1) faeces) were 9.55, 7.26, 7.01 and 6.93 for Lactobacillus, Enterobacteriaceae, coliforms and E. coli populations respectively. The Enterobacteriaceae, coliform and E. coli populations decreased through week 14 and 18, but were higher in animals fed with the DSM diet compared with the basal diet without DSM. The Lactobacillus populations at weeks 14 and 18 were lower in the animals fed the diet without DSM, whereas feeding DSM maintained the Lactobacillus counts from week 10. At week 22, populations of Enterobacteriaceae, coliforms and E. coli were >week 18 for the animals fed the diet without DSM, less change was observed with the feeding of DSM, and no differences between the diets were observed at week 22. However, in week 22 the animal gain was positively correlated with Lactobacillus numbers and negatively correlated with E. coli numbers. Subtraction of the E. coli population (log10) from the Lactobacillus population (log10) yielded a positive value termed 'effective'Lactobacillus that correlated well with animal gain and may better define a beneficial function in the intestine. Salmonella were detected in over 60% of the animals at week 10 and 14, and <20% at week 18 and 22. Campylobacter were detected rarely at weeks 10, 14 and 18, but were found in 25% of the animals at week 22. The DSM did not affect Salmonella or Campylobacter shedding, but examination of individual animals over the entire experiment indicated that fewer recurring incidences of Salmonella shedding occurred in animals that maintained higher Lactobacillus. In addition, at week 22, Salmonella and Campylobacter shedding was associated with lower levels of effective Lactobacillus and lower animal weight gains. CONCLUSIONS: The DSM did not directly affect the animal performance or pathogen shedding via the Lactobacillus spp. population at any phase of production. However, analysis of data from all animals revealed that faecal Lactobacillus affected Salmonella shedding and in the finishing phase, animal growth and pathogen shedding also were affected, as reflected by the 'effective'Lactobacillus-associated observations. SIGNIFICANCE AND IMPACT OF THE STUDY: In the swine intestine, any benefits from gastrointestinal Lactobacillus may be compromized by the E. coli population, and this antagonism may explain responses observed with prebiotics or probiotics in some swine.