OUP user menu

Ecology of Listeria monocytogenes in the environment of raw poultry meat and raw pork meat processing plants

Elise Chasseignaux, Pascale Gérault, Marie-Thérèse Toquin, Gilles Salvat, Pierre Colin, Gwennola Ermel
DOI: http://dx.doi.org/10.1111/j.1574-6968.2002.tb11192.x 271-275 First published online: 1 May 2002


The zoonotic Listeria monocytogenes is mainly transmitted to humans by the food-borne route. This bacterium was often found in the environment of food processing plants. Therefore the aims of this study were (i) the identification of environmental factors associated with L. monocytogenes contamination on working and non-working surfaces in poultry or pork processing plants and (ii) the understanding of its survival in such environments. The physicochemical risk profiles showed that a surface in resin or plastic, rather than uneven, with organic residues, with a neutral pH, a low temperature and a high hygrometry was associated with L. monocytogenes contamination.

Key words
  • Ecology
  • Poultry
  • Pork
  • Processing plant environment
  • Listeria monocytogenes

1 Introduction

Listeria monocytogenes has been described as one of the major human food-borne pathogens. Listeriosis can occur as a sporadic disease or as an outbreak and is often related to the consumption of contaminated food. In France, during the last 10 years, different outbreaks were associated with either delicatessen (pork tongue in jelly in 1992[1], pork ‘rillettes’ in 1993[2] and 1999–2000[3]) or soft cheeses (‘Brie de Meaux’ in 1995[4], ‘Pont l'Evêque’ and ‘Livarot’ in 1997, ‘Epoisses’ in 1999[3]). Different studies on L. monocytogenes prevalence showed that 16% of raw pork meat and 17% of raw poultry meat were contaminated[5]. The plant environment can also be contaminated: about 8% of samples in poultry slaughterhouses[6], 26% of samples in raw poultry meat plants[7] and 68% of samples in raw pork meat plants[8].

Therefore, the understanding of the survival of L. monocytogenes isolates is essential to prevent contamination in food plant environments. Actually, despite some studies on the tracing of L. monocytogenes in processing plants [912], to our knowledge the associated environmental risk factors of this microorganism are not completely analysed. Generally, ecological and physiological data came from laboratory experiments: L. monocytogenes could generally grow from 1 to 45°C[13], even if some strains can develop at 0.5°C[14] or at −0.2°C[15]. L. monocytogenes isolates could grow from pH 5.0 to 9.6; however, the optimal pH is neutral to slightly alkaline[16]. All these assays were realised with artificial media, therefore some nutriments or inhibitory substances found in natural environments could be missing.

The aim of this study was to identify the environmental risk criteria associated with L. monocytogenes colonisation on working and non-working surfaces in five processing plants: two of raw poultry (A and B) and three of raw pork meat (C, D and E). These surfaces were considered as the main source of meat contamination. The link between risk factors and physicochemical characteristics of the surfaces was studied by statistical analysis.

2 Materials and methods

2.1 Plants studied

Two poultry (A and B) and three pork (C, D and E) meat processing plants were studied. Plants A, B, C and D are in the North West of France whereas plant E is in the South East. The plants were divided into three different areas: reception of raw materials, meat processing and product processing. The working rooms were studied during the processing or after the cleaning operations at different times: 1 year (plant A), over 4 months (plant C), 2 months (plants B and D) and only one visit (plant E).

2.2 Swabbing

The sampling surfaces were grouped into two classes: (i) the environment for the surfaces without any contact with the raw meat (floor, wall, sewer) and (ii) the equipment for the surfaces in direct contact with the raw meat (working table, transport belt, knives, etc.). The swabbing was delimited with a sterile stainless steel frame (a surface of 162.5 cm2) using a tissue swab moistened with 5 ml tryptone salt, 10% (v/v) Isobio® (Laboratoire LCB, Lugny, France). Isobio® is used to neutralise cleaning products and disinfectants.

For each swabbing, different data were collected concerning the surface: composition (plastic, resin, stainless steel, metal, painted metal, tiling, cement or painted cement), cleanness (clean, presence of organic residues, presence of dust, presence of organic grease or encrustation), visual status (smooth, granular, stripped or damaged), moisture (dry surface, moisture or presence of water), temperature and pH of the surface. Surface temperature was measured using a thermocouple thermometer and the surface pH was determined using pH indicator paper. Room temperature and hygrometry were also collected using a digital hygrometer.

2.3 L. monocytogenes detection procedure

The tissue swab was resuspended in 90 ml half Fraser broth (bioMérieux, Marcy l'Etoile, France) and incubated at 30°C. After 24 h, each sample was streaked on Palcam agar (bioMérieux, Marcy l'Etoile, France) and incubated for 48 h at 37°C. The Vidas test (bioMérieux, Marcy l'Etoile, France) was then performed with typical colonies on Palcam agar plates. The plates were soaked with 3 ml tryptone salt broth and scraped. The obtained bacterial suspension was used to perform the Vidas test according to the manufacturer's recommendations and to our own validated protocol[17].

2.4 Statistical analysis

For each plant, an identical data analysis was performed. The dependent binary variable LIST was coded 1 if the swab was positive for L. monocytogenes; otherwise it was coded 0. Eleven independent environmental and physicochemical variables were listed. They were divided into two groups. The first group was composed of the variables for the room conditions (room, sampling surface, activity or cleaning operation) and the second group corresponded to the physicochemical and environmental conditions (composition, cleanness, visual state, moisture, temperature and pH of the surface, temperature and hygrometry of the room).

Frequencies were calculated for all the qualitative variables, and minima and maxima were noted. Histograms were drawn for all the quantitative variables and the mean, standard deviation, median and quartiles were calculated. Each quantitative variable was split into separate classes depending on our biological and microbiological knowledge. The class numbers were not under 10% of the swabs. Relationships between LIST and each of the other variables were then assessed in two-by-two tables and tested with the chi-square method. During this first screening, variables statistically associated with LIST (P≤0.15) were retained. Tests between independent variables were also performed so that redundant variables were eliminated.

Multiple correspondence analyses (MCA) were then performed using the software SPAD-N. The data of group 2 retained after the chi-square tests were active whereas the variables of group 1 retained after the second phase were illustrative. Successive MCA were performed and the variables with the highest inertia on the first axes were retained. Then the proximity between LIST+, LIST− and the modalities of the other variables were studied.

3 Results and discussion

3.1 Detection of L. monocytogenes

The examination of the 497 samples, of which 263 were realised during activity and 234 after the cleaning operations, showed that 23.7% of the samples were contaminated by L. monocytogenes. Table 1 indicates the contamination in the different plants either in the environment or on the equipment of the different studied rooms.

View this table:
Table 1

Distribution of the 497 swabbing samples regarding L. monocytogenes contamination in the five plants (A–E) in the environment or on the equipment of the different workrooms during activity and after cleaning

PlantType of swabbingNumber of samples contaminated by L. monocytogenes/total number of samples
ReceptionMeat processingProduct processingTotal
AEnviron.9/17 (53%)a11/27 (41%)2/5 (40%)0/7 (0%)7/14 (50%)0/6 (0%)18/36 (50%)11/40 (27.5%)
Equip.1/3 (33%)0/3 (0%)15/18 (83%)0/32 (0%)7/51 (13.7%)0/24 (0%)23/72 (31.9%)0/50 (0%)
BEnviron.4/4 (100%)ND8/12 (66.5%)2/8 (25%)3/8 (37.5%)2/8 (25%)15/24 (62.5%)4/16 (25%)
Equip.0/2 (0%)ND0/4 (0%)0/2 (0%)0/6 (0%)0/6 (0%)0/12 (0%)0/8 (0%)
CEnviron.1/6 (16.5%)0/5 (0%)0/11 (0%)0/11 (0%)0/6 (0%)0/5 (0%)1/23 (4.4%)0/21 (0%)
Equip.0/3 (0%)0/5 (0%)18/23 (78.2%)0/19 (0%)1/9 (11%)(0%) (7.1%)19/35 (54.3%)1/38 (2.6%)
DEnviron.0/4 (0%)0/4 (0%)1/3 (0%)0/3 (0%)2/4 (50%)0/4 (0%)2/11 (18.2%)0/11 (0%)
Equip.XbX4/11 (36.3%)0/11 (0%)5/8 (62.5%)0/9 (0%)10/20 (50%)0/20 (0%)
EEnviron.0/3 (0%)0/3 (0%)1/4 (25%)1/4 (25%)3/8 (37.5%)1/8 (12.5%)4/15 (26.5%)2/15 (13.3%)
Equip.1/1 (100%)0/1 (0%)7/7 (100%)0/6 (0%)0/7 (0%)0/7 (0%)8/15 (53.3%)0/15 (0%)
  • Act., during activity; Cle., after cleaning; Environ., environment (floor, wall, sewer); Equip., equipment (for example, working table, transport belt, etc.); ND, not determined.

  • aPercentage of positive samples.

  • bX, no equipment present in this room.

During processing, 38% of the samples contained L. monocytogenes, 38.9% in the poultry processing plants and 37% in the pork processing plants. This contamination was higher than that observed by Lawrence et al.[7] in a raw poultry meat processing plant (26%) but lower than that noticed by Salvat et al.[8] in a raw pork plant (55%). Nevertheless it was heterogeneous in the different plants. Two cases were beheld: an overall contamination either in the environment (plants A and B) or on the equipment (plants C, D and E). However, differences were observed when the different rooms were considered. In plant A, the environments of the reception and the product processing areas were the most contaminated (mostly on floors) whereas the contamination was present on the equipment of the meat processing area. In plant B, only the environments of all areas were contaminated (only on floors). In plants C and D, the environment of the reception area was colonised, and during meat and product processing the contamination was more important on the equipment (working tables, transport belts, etc.). However, in plant E, L. monocytogenes was detected on the equipment of the reception and meat processing areas and in the environment of the product processing workroom.

After the cleaning operations, the contamination was subsequently lowered (7.7%), with 13.1% in the poultry plants. This result is similar to those obtained by Salvat et al.[8]. In the pork plants, the contamination was slightly lower (2.5%). The residual contamination was mainly observed in the environments of plants A, B, and E: 27.5%, 25% and 13.3%, respectively, of contaminated swabs were found. On the contrary, in plant C, no contamination remained in the environment whereas one equipment was still contaminated. In plant D, no remaining contamination was detected in either kind of swabbing. This could be due to distinct cleaning habits: alternative utilisation of two different cleaning products whereas the other plants always used the same one. Therefore, in these other plants, some bacteria could develop a resistance against the used cleaning product, survive and eventually grow[18].

3.2 Profiles of risk factors for the environmental and physicochemical variables

Each plant was studied separately. Table 2 presents the characteristics of the surfaces and workrooms in the presence or absence of L. monocytogenes contamination. For some parameters, the results were not conclusive, meaning that no real difference was observed for the variable in either the presence or the absence of a contamination. However, the global analysis of the different parameters showed a convergence for some of them between the different plants whereas only inclinations or even divergences were observed in other cases.

View this table:
Table 2

Characteristics of the surfaces and workrooms in the absence or presence of L. monocytogenes contamination in the five plants studied

Absence of L. monocytogenes contamination
SurfaceNaturestainless steelstainless steelstainless steel; tilesNCcement; tiles
Cleannessclean or PDNCcleancleanclean or PD
Presence of L. monocytogenes contamination
  • NC, not conclusive; PD, presence of dust; POR, presence of organic residues; PORD, presence of organic residues and dust; PORG, presence of organic residues and grease. Results of the MCA: plant A (axis 1 (20.14%), axis 2 (16.49%), axis 3 (12.27%)), plant B (axis 1 (19.86%), axis 2 (17.94%), axis 3 (15.94%)), plant C (axis 1 (27.62%), axis 2 (15.32%), axis 3 (13.89%)), plant D (axis 1 (26.55%), axis 2 (21.64%), axis 3 (16.58%)), plant E (axis 1 (22.85%), axis 2 (19.06%), axis 3 (15.34%)).

Convergence was observed for the cleanness of the surface and the hygrometry of the room. A clean surface was associated with the absence of a contamination in almost all plants (A, C, D and E) even if the presence of dust was also recorded in plants A and E. However, in each plant organic residues were noticed on the surface in case of a L. monocytogenes detection.

In the absence of L. monocytogenes contamination, the hygrometry of the workrooms was below 70%, whereas in the presence of a contamination a higher hygrometry was observed (from 70 to 80% in plants B, C, E and even above 80% in plant A). However, no conclusion could be drawn for plant D, as during activity the hygrometry was always below 70% and conversely after the cleaning operations it was always above 80%. Helke et al.[19] observed a higher survival of L. monocytogenes in biofilm at 75.5% of relative humidity than at 32.5%.

Inclinations were observed for the status, the pH and the temperature of the surface and for the temperature of the workrooms. A smooth surface was clearly linked with the absence of L. monocytogenes detection. However, when L. monocytogenes was detected, the surface was uneven: granular (plants A and B), stripped (plant D) or damaged (plant C).

A pH under 6 was frequently associated with the absence of L. monocytogenes. However, when L. monocytogenes was detected, differences in pH were noted but nevertheless close to neutral (a pH from 6 to 6.5 was linked with the contamination in plants C and E, whereas for plant A it was from 6.5 to 8, and for plant B above 6.5). This is in agreement with the growth characteristics of L. monocytogenes, whose optimal pH is neutral to slightly alkaline[16].

Regarding the temperature of the surface, differences were observed. In the absence of contamination, the surface temperature seemed rather high, above 10°C (above 4°C in plant D, between 10°C and 13°C in plant A, above 12°C in plant E and above 20°C in plant B). The only exception was observed in plant C where the temperature was either below 8°C or above 13°C. On the contrary, when L. monocytogenes was detected, the temperature was low, below 10°C. However, in plant C the temperature was between 10°C and 13°C but the contamination was still associated with the refrigerated temperature. The same results were obtained for the room temperatures: the temperature was rather high in the absence of L. monocytogenes contamination and inversely low in the presence of a contamination. This result was expected as the surface temperature was partly influenced by the room temperature. This was in agreement with the psychrotrophic properties of L. monocytogenes: this ability could select L. monocytogenes while other competitive microflora could not grow [16,20]. On the contrary, at high temperature, the microflora could compete with L. monocytogenes, outnumber it, and therefore not allow its implantation. Moreover, our study confirmed the results of Helke et al.[19], which showed a higher survival when the biofilm was at 6°C compared to 25°C with a relative humidity of 75.5%. In our study, L. monocytogenes contamination was found with low temperature and high hygrometry and, conversely, no contamination with high temperature and low hygrometry.

Finally, for the nature and the moisture of the surface, divergences were observed. When no L. monocytogenes was detected, the surface was of stainless steel (plants A, B and C), tiles (plants C and E) or cement (plant E). When a contamination was observed, the surface was composed of either resin (plants A and B) or plastic (plants C and E). Nevertheless, this was in agreement with the surface state: smooth in the absence of L. monocytogenes detection and granular, stripped or damaged with L. monocytogenes contamination. Moreover, in the different studied plants, the resin was used on the floor as a non-skid surface and therefore it was granular. On plastics, pits and cracks resulted of their use. Thus, for both surface kinds, microspaces existed, unreached by disinfectants, where soil and bacteria could persist. Wong[21] also supported that hypothesis. Moreover, different studies related to biofilms on stainless steel showed that nutrients could either enhance or inhibit a biofilm of L. monocytogenes[19,22,23]. In our study, this surface was found smooth or lightly stripped and associated with the absence of contamination by L. monocytogenes. Therefore, organic meat residues could limit the establishment of L. monocytogenes biofilm on stainless steel.

Concerning the moisture of the surface, a dry surface was found in three plants (A, B and E) whereas a moist surface was found in two plants (C and D) with no detection of L. monocytogenes. Opposite results were noticed when L. monocytogenes was detected. Therefore, no conclusion can be expressed for that criterion.

As a conclusion, the physicochemical risk profiles showed that a surface in resin or plastic, therefore rather uneven, with organic residues, with a neutral pH, a low temperature and a high hygrometry was associated with L. monocytogenes contamination. Now, it could be interesting to consider the accompanying flora found in either the presence or the absence of L. monocytogenes contamination. Indeed, as well as physicochemical factors, flora could also influence the surface colonisation by L. monocytogenes.


The authors acknowledge the Ultra-propre Nutrition Industrie Recherche group for financial support for the investigations, the French Ministère de l'Education Nationale, de la Recherche et de la Technologie and the Agence Nationale pour la Recherche et la Technologie (ANRT) for its grant. The authors also acknowledge Y. Le Nétre-Michel, F. Eono and S. Gorin for their technical help during the study.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
View Abstract