Microbiological Analysis of Wild Lowbush Blueberries Harvested in Nova Scotia, Canada for the Fresh Produce Market

Canada is a leading producer of wild lowbush blueberries, most of which are mechanically harvested, washed, individually quick frozen (IQF), and bulk packaged. Still, some berries are harvested by more gentle methods and sold as fresh-packed products. These berries do not undergo a wash step, nor are subjected to antimicrobial treatments. The purpose of this study was to conduct a microbiological survey of berries harvested in the province of Nova Scotia to assess their potential for harborage of bacterial foodborne pathogens. A combination of standardized plate count methods and 3M-Petrifilm protocols were used to enumerate total aerobic mesophilic bacteria (APC), yeasts and molds (YMC), coliforms, and generic <i>E. coli</i>, the latter being an indicator of fecal contamination. Overall, APC and YMC levels were 1.2 and 0.5 log greater, respectively, for berries collected early in the harvest season versus those acquired late season and varied significantly (<i>p</i> < 0.05) between farm (location) and harvest practices used. Berries harvested by our team using sanitized hand rakes (SH) had consistently lower APC and YMC levels than those harvested by farm crews. Yet, when gentle harvesting (GH) methods (hand-raking, walk-behind or modified mechanical harvesters) were employed on farms, lower numbers were generally observed compared to berries harvested by traditional tractor-mounted mechanized harvesters (MH). The presence of coliforms (and their levels) was also impacted by the harvest method, with detection rates of ~29%, 73%, and 92% in SH, GH, and MH samples, respectively. Mean counts were < 2.5 log₁₀ CFU/g for both SH and GH berries, but significantly higher (<i>p</i> < 0.05) on MH berries (3.6 log₁₀ CFU/g). Although ~56% of all berry samples collected (n = 350) contained coliforms, only 12 were positive for <i>E. coli</i>, 9 of which were MH samples. Only the latter had numbers > 2 log₁₀ CFU/g, but none tested positive for Shiga toxin-producing serotype O157 (STEC O157) or <i>Salmonella</i> spp. when using internationally recognized selective enrichment and plating methods. ATP luminescence was used to assess the general hygiene of processing lines, whereby "hot spots" for microbial activity were identified, even after cleaning., Standard selective enrichment and plating methods were used for the detection of <i>Listeria monocytogenes</i> on 61 swab samples taken from berry totes or conveyor belts at different times during processing; 4 swabs tested positive for <i>L. monocytogenes</i>. However, the pathogen could not be detected by direct plating on selective agar without prior enrichment; this indicated its numbers were low. The results from this work demonstrated that alternative gentle harvest methods can reduce microbial numbers on wild blueberries. Although the frequency of fecal contamination in berry samples appeared to be low and targeted human pathogens were not detected; this represents a single study conducted over one harvest season. Therefore, it would be prudent for processors to seek effective antimicrobial technologies prior to packaging, while consumers should use caution and thoroughly wash produce before consumption. Where sporadic detection of <i>L. monocytogenes</i> was observed on environmental samples from the processing line, processors must ensure that effective sanitation programs are implemented to avoid potential food safety risks.