Environmental Screening of a Cannabis Production and Processing Facility: A Comparison of an Environmental DNA Microarray and Traditional Microbiological Plating Methods

December 18, 2019

A case study that demonstrates the utility and necessity of environmental screening in a cannabis production and processing facility.

As the cannabis industry continues to expand and become more heavily regulated, the need for screening tools that detect microbial contamination increases. While screening has primarily focused on the raw product, there has been little emphasis on the actual facilities in which that product is processed, which has the potential to be a contaminating source for the cannabis product. The following case study was performed to demonstrate the utility and necessity of environmental screening in a cannabis production and processing facility. Samples were collected for assessment of microbial contamination across 11 locations throughout the facility. Each sample was assessed by traditional microbiological plating and an environmental DNA microarray to compare the effectiveness of both methods.

Environmental surveillance is increasingly appreciated for its utility in public health efforts, including those related to agriculture and quality control (1). The presence of pathogenic organisms in production facilities is concerning because it indicates that there may be a reservoir of the contaminating organisms within the facility (2). Importantly, these pathogenic organisms may be capable of posing a significant risk to both human health and agricultural products. More specifically, the presence of agricultural pathogens in horticultural operations may increase the risk of agricultural disease, thereby increasing the likelihood of agricultural losses and rejected batches of product. These concerns persist not only in agricultural operations, but also in food and drug production.

Provided the potential ramifications of microbial contamination in such a facility, it is important that practices in risk reduction are diligently followed and maintained. An important component of risk reduction is the continuous screening and monitoring of a facility for microbial contamination. Importantly, while an initial investigation may be able to detect and identify contaminating organisms present within a facility, subsequent screening and surveillance is necessary to confirm the effectiveness of decontamination methods, verify sources of contamination, and validate the cleanliness of a facility, as well as identifying any new instances of contamination that may occur.

Currently, the golden standard for the screening, detection, and assessment of microbial contamination has been through culturing on agar plates (3). While traditional plating methods offer a visual confirmation of microbial presence and viability, and do not require sophisticated equipment, this method does pose several disadvantages. For instance, plating is incredibly time-consuming, can be laborious, and requires the expertise of a skilled microbiologist. Additionally, a comprehensive microbiological analysis, which encompasses the detection of multiple types of organisms, often requires the use of multiple plates and media types, increasing material and labor costs; even so, species-level identifications require secondary methods of confirmatory testing.

Though their use has not yet become standard in the industry, the use of microarrays for broad-spectrum microbial detection and identification has been well-studied for its many applications in clinical and agricultural settings (4). Microarray technology is able to identify microorganisms at the family, genus, and species-level of classification. Multiplexed in design, microarrays are capable of simultaneously detecting a multitude of microbial isolates within a single sample in a matter of a few hours, making the array more cost-effective and less labor-intensive than traditional plating methods. While microarrays do require specialized equipment and training in molecular techniques, the mechanism of recognition of specific DNA sequences imparts a high degree of both specificity and sensitivity in microbial detection.

This study was conducted in an effort to highlight the utility and necessity of environmental screening of an agricultural production and processing facility. The study also compares the proficiency of traditional microbiological plating and the environmental DNA microarray used in this screening.


Sample Collection
First, 56 swabs (each suspended in 4 mL of buffered peptone water) were provided to the collaborating facility for this study. Sampling locations and intervals of sample collection were determined at the discretion of the facility contact. Immediately following sample collection (performed according to the procedure outlined in PathogenDx Product Insert: EnviroX Environmental Swab), swabs were stored at 4 °C until return shipment to PathogenDx (samples were shipped with ice packs to ensure samples remained cold over the course of delivery). All swabbing was conducted between March 7, 2019 and April 15, 2019, and samples were returned to PathogenDx over the course of three shipments shortly after collection time.

Sample collections spanned across 11 distinct rooms within the facility:

  • Veg 1

  • Veg 2

  • Propagation

  • Post-Harvest

  • Dry Room

  • Mother Room

  • Clone Room

  • 3rd Party Lab Sampling Room

  • Packaging

  • Packaging 2nd Room

  • Inventory

Documentation was provided for each sample collection, including date of collection, site of collection, and whether the swab site was swabbed prior to or after decontamination procedures were performed (“dirty” and “clean” designations, respectively). Schematics were also provided for seven of the above locations (Veg 1, Veg 2, Propagation, Post-Harvest, Dry Room, Mother Room, and the Clone Room).

Sample Analysis
Upon arrival, swab collections were homogenized by vortexing and aliquoted into 1 mL samples for analysis via traditional microbiological plating and PathogenDx EnviroX.

Traditional Microbiological Plating
Each sample was plated on a general medium for bacteria (Tryptic Soy Agar, TSA) and fungi (Sabouraud Dextrose Agar with chloramphenicol, SDA) to capture as many microbial contaminants as possible.

In addition to a neat sample, a 1:10 and 1:100 dilution was plated for each swab collection to ensure that individual isolates could be visualized in the event that heavy concentrations of microbial contaminants within the samples resulted in overgrowth of the plates. For each concentration, 100 µL of sample was plated.

Both plate types were incubated at room temperature (25 °C); TSA plates were incubated for two days and SDA plates were incubated for six days.

PathogenDx EnviroX Microarray
Microarray analysis was performed on 1 mL of the original sample according to the procedure outlined in the PathogenDx Product Insert: EnviroX Environmental Swab.


For each sample, the results of the traditional plating and EnviroX methods were compared directly; a representative example of the comparative analysis is demonstrated in Figure 1, which displays the results for swab number one (located in Veg 2 within the facility). Microbial contamination is apparent, and robust, on both TSA and SDA plates, representing bacterial and fungal growth, respectively. While a dilution effect can be observed in the growth on the SDA plates, the growth observed on the TSA plates remained too concentrated to discern individual colonies. Interestingly, swab number one was collected after cleaning and disinfectant procedures had been conducted for this particular swab site (“clean” designation). While unable to definitively identify the contaminants from the plates alone, several categorical and species-level identifications were ascertained by the EnviroX microarray, including total aerobic bacteria (TAB), total enterobacteriaceae (TE), bile-tolerant gram-negative (BTGN), and total yeast and mold (TYM); species-level identifications included Aeromonas spp., Pseudomonas spp., and Cladosporium spp. Similar trends were observed throughout the remaining study samples.

Of the 56 swab collections analyzed, all samples tested positive for microbial contamination through both traditional microbiological plating and EnviroX analysis methods (Table I). (See upper right for Table I [in two parts as Table 1.1. and 1.2], click to enlarge.) In comparing the number of swabs with confirmed growth on TSA (bacterial) to the number of confirmed detections on TAB (bacterial) on the EnviroX array, EnviroX displayed equal or greater sensitivity in detecting contaminating microorganisms as compared to the microbiological plating in all cases (Table II). The same trend was observed, to a greater extent, in comparing the number of swabs with confirmed growth on SDA (fungal) as compared to the number of swabs with confirmed detection of TYM (fungal). Notably, TE and BTGN are more defined subcategories of TAB and consequently, were detected at a lower frequency than TAB.

The composition of microbial species detected was variable depending on the location analyzed (Table III). Veg 1 presented with the highest degree of variability with the detection of seven species-level identifications (Aeromonas spp., Pseudomonas spp., Pseudomonas aeruginosa, Fusarium oxysporum, Candida spp., Penicillium spp., and Mucor spp.).  In contrast, the Clone Room presented with the lowest degree of variability with the detection of only one species-level identification (Pseudomonas spp.).

Given the high degree of consistency in organisms detected across different locations within the facility, the frequency of locations that tested positive for each organism was calculated (Table IV). Pseudomonas spp. appeared to be the most prevalent microbial species within the facility, appearing in 82% of the locations tested (9 out of 11 locations). The next most prevalent species found was Golovinomyces spp., and was observed in 55% of the locations tested (6 out of 11 locations). Many microbial species observed were isolated to singular locations within the tested locations (Pseudomonas aeruginosa, Mucor spp., Aspergillus terreus, and Aspergillus fumigatus).

Provided the species-specific identifications from the EnviroX analysis, temporal and spatial relationships in the distribution of microbial contaminants were evaluated. An observed temporal relationship is observed in Figure 2, which displays the contamination present at two sites within Veg 1, before (“dirty”) and after (“clean”) decontamination procedures were utilized. Looking at the two sets of agar plates, there is evident reduction of microbial burden after decontamination procedures were utilized, but they were not sufficient to remove all microbial contamination at these two sites. Interestingly, the microarray data indicated that four species of microbial contaminants were present before decontamination (Aeromonas spp., Pseudomonas spp., Fusarium oxysporum, and Candida spp.). After decontamination, only Candida had been removed from the sites, and the other three species still remained.

Among others, spatial relationships were observed in the distribution of Golovinomyces and Cladosporium spp. throughout the facility. As demonstrated in Table IV, Pseudomonas spp. was detected in 82% of the locations swabbed, and was the most widely distributed organism within the facility. Golovinomyces spp. was the second most prevalent, appearing in 6 of the 11 locations swabbed. When broken down by location, it was observed that Golovinomyces spp. was detected in 86% of the swab sites within the Post-Harvest room (7 of 8 swab sites). In addition to the Post-Harvest room, Golovinomyces spp. was identified in the Dry Room, 3rd Party Lab Sampling room, Packaging, Packaging 2nd room, and Inventory. Followed closely behind, Cladosporium spp. was present in 5 of the 11 locations swabbed, and was primarily concentrated in Veg 2, where all 9 of the swab sites within the room tested positive (Figure 3).


This study highlights the utility of environmental screening as a tool to evaluate potential microbial contamination within an agricultural production and processing facility. While largely harmless, many of the microorganisms detected in this surveillance carry potential risks to both human health and agricultural products and yields.

Highly ubiquitous in the environment and agricultural samples, it is unsurprising that Pseudomonas spp. was the most prevalent isolate in this study; while some species of Pseudomonas can be harmful, many serve as plant commensals and are not especially concerning (5). Conversely, the detection of organisms such as Golovinomyces spp. and Cladosporium spp. are more concerning. The presence of Golovinomyces spp. is particularly problematic from an agricultural perspective because these species are a major cause of powdery mildew in plants, which is capable of reducing or destroying agricultural yields (6). By comparison, Cladosporium spp. is relatively ubiquitous in the air, but can be a significant allergen and can pose health concerns in susceptible individuals (7). Without environmental screening, microbial contamination such as this may go undetected, imposing risks to agricultural yields and human health. Provided such screenings, steps can be taken to reduce contamination, modify decontamination procedures as necessary, and monitor facilities to ensure rapid detection of any recurrence of contamination.

Further, this study emphasizes the advantages of utilizing the PathogenDx EnviroX microarray technology in microbial detection as compared to traditional microbiological plating. In addition to producing results in a more cost-effective and rapid manner as compared to traditional plating, the EnviroX microarray displayed equal or greater sensitivity in detecting microbial contamination in all sample cases (Table II). In fact, there were many cases in which the agar plates displayed no growth but EnviroX detected contamination, for both bacterial and fungal isolates. In addition, EnviroX provided speciation of many of the contaminants present, a distinguishing characteristic that could not be determined by the agar plating alone. Notably, without these species-level identifications, the observed temporal and spatial relationships could not have been ascertained. While some patterns can be observed from the agar plates, the species-level identifications could not be made without further experimental analysis. Furthermore, without species-level identifications, the degree of risk associated with the specific contaminants present cannot be fully appreciated. Taken together, these data support the usefulness and need for environmental screening in agricultural processing facilities, and highlights the critical advantages in utilizing microarray technology for microbial detection, as opposed to traditional microbiological methods.


  1. S.L. Groseclose and D.L. Buckeridge,  Annu. Rev. Public Health38, 57–79 doi:10.1146/annurev-publhealth-031816-044348 (2017).
  2. F.E. Bartz, J.S. Lickness, N. Heredia, F. Fabiszewski de Aceituno, K.L. Newman, and J.S. Leon, Appl. Environ. Microbiol. 83(11), e02984-16 (2017).
  3. H.M. Davey, Appl. Environ. Microbiol.77(16), 5571–5576 doi:10.1128/AEM.00744-11 (2011).
  4. T. Kostic and A. Sessitsch, Microarrays 1, 3–24 doi:10.3390/microarrays1010003 (2012).
  5. R. Sitaraman, Front. Plant Sci.6(787) doi:10.3389/fpls.2015.00787 (2015).
  6. A. Lebeda and B. Mieslerova, Plant Pathol.60(3), 400–415 doi:10.1111/j.1365-3059.2010.02399.x (2010).
  7. A. Bozek and K. Pyrkosz, Hum. Vaccines Immunother.13(10), 2397–2401 doi:10.1080/21645515.2017.1314404 (2017).

About the Authors

Chelsea Adamson is a microbiologist and Benjamin A. Katchman is a principal scientist at PathogenDX in Tucson, Arizona. Direct correspondence to: bkatchman@pathogendx.com

How to Cite this Article

C. Adamson and B.A. Katchman, Cannabis Science and Technology2(6), 54-61 (2019).

download issueDownload Issue : November/December 2019