Avaliação do perfil microbiológico do ar ambiente da área limpa de um abatedouro frigorífico de suínos

Abstract

The evaluation of the microbiological quality of air in industrial environments is essential to ensure food safety. This study was conducted in the clean area of a swine slaughterhouse located in the western region of Paraná, Brazil, with the objective of monitoring the microbial load in the air throughout the processing period. The absence of specific regulations for microbiological air analysis in food industries in Brazil highlights the relevance of this research. Indicator microorganisms, such as aerobic mesophilic bacteria, molds, and yeasts, were used in the analyses. Air samples were collected by sedimentation at five strategic points within the clean area, at six different time intervals throughout the operational shift. After incubation, the counts were expressed as CFU/plate/incubation time. Additionally, ambient temperatures at the sampling points were recorded to assess the influence of temperature on microbial growth. The data obtained were organized into spreadsheets and analyzed through nonlinear regression using Statistica® software, with subsequent fitting to the mathematical models of Brody, Von Bertalanffy, Richards, and Gompertz. The results showed significant variations in microbial load between the analyzed points and times. Points P3 and P4, located at the viscera table and the evisceration platform, respectively, exhibited the highest contamination levels. In contrast, point P2, situated near the final wash shower and the cooling antechamber, showed the lowest counts in both analyses, likely due to the effect of residual chlorine and the lower temperature recorded at that location. In the statistical analysis, the Brody model performed best in describing the growth of aerobic mesophilic bacteria (R² = 0.9999), while the Von Bertalanffy model was more suitable for representing the growth of molds and yeasts (R² = 0.9886). The combined evaluation of statistical parameters, such as mean relative error, root mean square error, and the sum of squared errors, confirmed the satisfactory fit of the models. It is concluded that all tested models were adequate to represent the experimental data, with Brody being the most suitable for aerobic mesophilic bacteria and Von Bertalanffy for molds and yeasts. The results obtained contribute to understanding the microbiological behavior of air in industrial environments and provide valuable insights for improving hygienic-sanitary control in processing areas.