Purpose: The aim of this study was to evaluate the prevalence of pathogenic and indicator microorganisms on broccoli fertilized with pig slurry or mineral fertilizer and irrigated 0, 1 or 2 times with contaminated water.
Methods: Twenty-four broccoli experimental plots were set using a split-plot factorial design with the fertilizer as the main plot factor and the number of irrigations as the subplot factor. Irrigation water was artificially contaminated with bovine slurry. Water samples were taken during irrigation, and broccolis were sampled 1, 3 and 5 days after irrigation. Soil samples were also taken on the last sampling day. Samples were analyzed to determine generic E. coli populations, and presence of generic E. coli, Salmonella spp., Listeria monocytogenes, and verotoxigenic E. coli. A generalized linear mixed model using SAS GLIMMIX procedure was performed to evaluate the impact of the treatments on E. coli prevalence.
Results: Generic E. coli populations in irrigation water were 393 and 538 CFU/100 ml for the first and second irrigation, respectively. Salmonella spp. was detected in pig and bovine slurries but no pathogenic microorganisms were recovered in water, broccoli or soil. Eleven broccoli samples were E. coli-positive using the enrichment procedure (n=72). None came from non-irrigated plots. E. coli prevalence on slurry-fertilized broccoli was 25 and 33 % when 1 and 2 irrigations were performed, respectively (n=12). E. coli prevalence on mineral-fertilized broccoli was 17 % for both irrigation treatments (n=12). However, statistical analysis showed no impact of treatments on E. coli prevalence. All soil samples taken in slurry-fertilized plots were E. coli-positive (n=12), while 9 out of 12 soil samples were positive in mineral-fertilized plots.
Significance: This study highlights the limited prevalence of E. coli on broccoli after irrigation, even if bacteria were frequently detected in soil.