Modelling the effect of oxygen concentration on bacterial growth rates.

Predicting the microbial safety of food products stored in modified atmosphere packaging implies taking into account the effect of oxygen reduction on microbial growth. According to their respiratory-type, the micro-organisms are not impacted similarly by the oxygen concentration. The aim of this article was to quantify and model the oxygen effect on the growth rates of 5 bacterial species: Listeria monocytogenes and Bacillus weihenstephanensis (facultative anaerobic), Pseudomonas fluorescens (strict aerobic), Clostridium perfringens and Clostridium sporogenes (strict anaerobic). The results showed the oxygen concentration doesn't modify the behavior of both facultative anaerobic strains. The growth rate of P. fluorescens decreased with the oxygen concentration, but the effect is only noticeable when the oxygen concentration fell below 3% in the gaseous phase. Conversely, the oxygen acted as a growth inhibitor for both Clostridium species. But total inhibition is reached only for 3.26% and 6.61% respectively for C. sporogenes and C. perfringens. Two models have been fitted for both respiratory-types, the first is the Monod model considering oxygen as a substrate for growth, and the second is the classic inhibitory model based on minimal inhibitory concentration.

Tracking spore-forming bacteria in food: from natural biodiversity to selection by processes.

Sporeforming bacteria are ubiquitous in the environment and exhibit a wide range of diversity leading to their natural prevalence in foodstuff. The state of the art of sporeformer prevalence in ingredients and food was investigated using a multiparametric PCR-based tool that enables simultaneous detection and identification of various genera and species mostly encountered in food, i.e., Alicyclobacillus, Anoxybacillus flavithermus, Bacillus, B. cereus group, B. licheniformis, B. pumilus, B. sporothermodurans, B. subtilis, Brevibacillus laterosporus, Clostridium, Geobacillus stearothermophilus, Moorella and Paenibacillus species. In addition, 16S rDNA sequencing was used to extend identification to other possibly present contaminants. A total of 90 food products, with or without visible trace of spoilage were analysed, i.e., 30 egg-based products, 30 milk and dairy products and 30 canned food and ingredients. Results indicated that most samples contained one or several of the targeted genera and species. For all three tested food categories, 30 to 40% of products were contaminated with both Bacillus and Clostridium. The percentage of contaminations associated with Clostridium or Bacillus represented 100% in raw materials, 72% in dehydrated ingredients and 80% in processed foods. In the last two product types, additional thermophilic contaminants were identified (A. flavithermus, Geobacillus spp., Thermoanaerobacterium spp. and Moorella spp.). These results suggest that selection, and therefore the observed (re)-emergence of unexpected sporeforming contaminants in food might be favoured by the use of given food ingredients and food processing technologies.