Comparative performance of air sampling methods in a broiler house: implications for use and recovery of bacteria

Accurate quantification of airborne bacteria in broiler houses is essential for assessing exposure risks to workers and animals, as well as for developing informed sanitation and surveillance strategies. Numerous sampling methods and equipment with differing collection physics, sampling parameters, processing, and handling have been developed in order to collect and quantify bacteria within bioaerosols. The objective of this study was to conduct a field comparison of three commonly used air sampling methods, including a cyclonic liquid impinger, a single-stage impactor, and gelatin filters housed in two-piece cassettes. Sampling was performed in an active broiler house using a balanced design (n = 6, k = 3, days = 5, N = 90) across five days (broiler ages D37, D38, D39, D40, and D44), with variation across sampling days evaluated descriptively. Liquid impinger and gelatin filter samples were collected from 100 L of air, whereas impactor samples were collected from 10 L to maintain countable colony densities. All counts were standardized to log₁₀ cfu/m³ before statistical analysis. Sampler type was found to have a statistically significant effect on bacteria recovery (P < 0.05), with method effects evaluated independently for total aerobic bacteria and Staphylococcus spp. cultured on tryptic soy agar and mannitol salt agar using a linear mixed-effects model that accounted for day level and block level variability. MacConkey agar results yielded frequent non-detections, with only sporadic recoveries on 100 L impactor samples, indicating that Gram-negative bacteria including coliforms were present at concentrations near or below method specific limits of detection. Although differences between samplers were modest compared to variability between days (accounting for 34 to 51% of total estimated variance across media), the impactor consistently yielded the highest recoveries across both tryptic soy agar and mannitol salt agar (4.47 and 4.18 log₁₀ cfu/m³, respectively), with recoveries from the liquid impinger and gelatin filters overlapping across media. These results indicate that sampler choice and associated processing can produce statistically significant but modest shifts in culturable airborne bacterial estimates that should be treated as a controlled methodological factor when interpreting and comparing field data. However, given the relative comparability between methods, ease of use and intended sampling parameters should guide sampling method selection rather than expectations of differing bacterial recovery.