To determine the prevalence of endophytic fungi and yeasts, greenhouse-cultivated cannabis plants at various stages of development were studied. Mother plants (varying in age from 2 months to 6 months) (Figure 1a), cuttings derived from them (Figure 1b), and vegetative (4-week-old) (Figure 1c) and flowering plants (4-8 weeks old) (Figure 1d) grown as described elsewhere [8,81] were obtained from up to 10 cannabis genotypes for microbiome analysis. None of these plants displayed any symptoms of disease. Root, stem, and leaf samples were collected randomly from these plants and sterilized using 70% ethanol and 10% bleach (containing 0.525% NaOCl) in various combinations of time, ranging from 30 sec to 3 min, depending on the tissue source. Also, three seed samples were included for analysis: hemp seeds from a supplier in Colorado, high-THC cannabis seeds from a greenhouse producer in British Columbia, and seeds from an online supplier. After tissue samples were harvested, they were flash-frozen in liquid nitrogen and stored at −80 °C. Samples were homogenized into a fine powder using a mortar and pestle under liquid nitrogen to ensure thorough cell disruption. Genomic DNA was extracted using the Qiagen DNeasy Plant Mini Kit (Qiagen, Inc., Toronto, ON, Canada) following manufacturer’s protocol, with additional steps to optimize the yield and purity tissues containing high levels of secondary metabolites. DNA quality and quantity were assessed using a NanoDrop spectrophotometer, while DNA integrity was confirmed through electrophoresis on a 1% agarose gel. DNA samples were sent to Medicinal Genomics (Beverly, MA, USA) (https://medicinalgenomics.com/genomic-services/ (accessed on 10 March 2024)) for high-throughput metagenomic sequencing (Next Generation Sequencing). Both fungal and yeast communities were profiled to capture the diversity of these microbial communities. Approximately 60 samples were assayed in total. The data from the same tissue types were combined for analysis. Sequencing data were saved as FASTA files containing adapter-trimmed high-quality reads. Reads were classified using a metagenomics classification algorithm that assigned a taxonomic identification to each sequencing read and estimated the abundance of each species in the sample. Taxonomic outputs were visualized using Krona plots to enable interactive and hierarchical visualization of microbial diversity. This approach facilitated a detailed exploration of microbial community structure and differences across plant parts and growth stages. The data presented in Figure 2 represent the percentage of total reads of the specific genus relative to total fungal reads in the tissue samples from specific plant growth stages. To confirm the presence of certain culturable endophytes, surface-sterilized stem pieces were incubated on potato dextrose agar containing 130 mg/L of streptomycin sulfate for up to 2 weeks to recover fungal endophytes [17].

The results from the microbiome analysis of all cannabis tissue sources are presented for fungal genera in Figure 2. Only the genus names are presented except for select species that are discussed in more detail below. These microbes represent reads that were >1% of the total reads for each microbial category. Genera with significantly higher reads are also identified in the figures. For fungal endophytes, 20 genera are shown in Figure 2, with the most abundant being (in decreasing order) Fusarium, Penicillium, Rhizophagus, and Aspergillus. These genera were present in all tissue types and plant sources, including seeds (Figure 2). The data for the presence of bacterial endophytic species are not presented.

The remaining genera present in all tissue types at a lower frequency (ca. 1%) included Alternaria, Botrytis, Colletotrichum, Phycomyces, Pyricularia, and Trichoderma. Other genera that were present in the roots of flowering plants were Beauveria and Metarhizium, which are biocontrol agents that were applied during cultivation and likely became established in the treated plants. While the plant pathogens Rhizoctonia, Mycosphaerella, and Verticillium were present in the growing medium, they were detected infrequently in host tissues. Finally, the genus Thielavia, a heat-tolerant saprophytic fungus capable of degrading cellulytic substrates [92], was present in the growing medium but was only detected in the roots of one sample.

Specific fungal genera present in the growing substrate were strongly correlated with those found in cannabis tissues, suggesting that this was a likely source of many of the fungi. These included Fusarium, Penicillium, Aspergillus, Colletotrichum, Pyricularia, and Trichoderma (Figure 2). Of note is that many of these fungi are plant pathogens. However, some fungi present in plant tissues but not present in the substrate, such as Alternaria, Beauveria, Botrytis, Peziza, and Phycomyces, likely originated from external environmental sources, as well as from the air, and subsequently became established in the plants. Previous air sampling studies conducted in greenhouses have shown that many of these fungi produce air-borne spores and can be detected on agar settling plates [17,32,81]. Exposed pruning sites and cut surfaces on stems are prone to colonization by various fungi in cannabis cultivation facilities [17,81]. Other genera that were found sporadically in the tissues included Chaetomium, Mycosphaerella, Mucor, Peziza, and Sordaria. These results demonstrate the dynamic nature of fungal endophyte establishment in cannabis plants under greenhouse conditions that can originate both from the growing substrate and from various external sources. Many of these fungi cannot be considered as beneficial because they include plant pathogens, and others are designated as saprophytes. In addition, the findings indicate that a range of endophytic fungi are inherently present in mother plants, and they can be carried over in cuttings taken from these plants which are used for vegetative propagation to initiate a new cycle of crop production. Such horizontal transmission indicates that resident microbes in cannabis cuttings originating from mother plants or seeds may already contain a resident microbial background (load). This should be considered when evaluating the effects of externally applied microbes to determine the impact on plant growth, as the potential influence of existing fungal (and bacterial) species already present internally should be recognized. The analysis of a limited number of seed samples showed that they may contain up to 12 genera of fungi. These included Alternaria, Botrytis, Colletotrichum, Fusarium, Mycosphaerella, Mucor, Pyricularia, and Phycomyces. The source of these fungi cannot be ascertained as the seeds were obtained as samples that did not originate from the plants analyzed in this study. Seeds are known to harbor a range of endophytic fungal and bacterial species [21,33,35].

To potentially reduce the background microbial load within a commercial cultivation operation, using plants initiated from meristem tip cultures is a proven approach [88]. In addition, the experimental application of a systemic fungicide, such as ‘Luna Privilege’ (active ingredient fluopyram), to mother plants strictly in a research laboratory setting was previously shown to promote the success rate of propagation in tissue cultures by eliminating the most prevalent endophyte contaminants [88]. These included species of Penicillium, Aspergillus, Fusarium, Chaetomium, and other fungi that are found in cannabis plants. While fungicides such as ‘Luna Privilege’ are not currently registered for use on cannabis, our previous observations suggest that reducing the existing endophytes can allow for the studies of plant performance following the application of supplemental endophytic species to be more meaningful. Similarly, seed treatments that include bleach or fungicides should reduce the frequency of seed-borne endophytes and epiphytes. The use of heat treatment of seeds or potentially of cannabis cuttings may also be approaches to reduce the carry-over effects of detrimental endophytes.

Surface-sterilized stem and leaf tissues from mother plants and flowering plants of cannabis incubated on an agar medium containing streptomycin sulfate (to inhibit the growth of bacteria) confirmed the presence of a range of endophytic fungal genera in these tissues (Figure 3). These included Aspergillus and Penicillium (Figure 3a), Fusarium (Figure 3b), Penicillium (Figure 3c,d), yeasts (Figure 3e), bacteria (Figure 3f), and a mix of Aspergillus, Penicillium, and other unidentified fungi (Figure 3g,h). The recovery of culturable fungi correlated well with the results from the microbiome analysis (Figure 2). Culturing methods were previously used to isolate and identify these species from various cannabis tissues [17,32,81].

An interesting finding in the present study was the ubiquitous presence of Penicillium chrysogenum in all 60 samples, including seeds, at detectable levels. In addition, Rhizophagus irregularis (former name Glomus intraradices) [93], an arbuscular mycorrhizal fungus, was also present in all tissue samples, including seeds (Figure 2). Penicillium chrysogenum produces the antibiotic penicillin [94], and its presence in all cannabis tissues of varying genotypes suggests that it has become adapted to this host and may be playing an important role, although additional studies are required to confirm this. This species was reported in a previous study to be present on cannabis inflorescences as a contaminant during the drying process [32]. A colony of P. chrysogenum isolated from cannabis stems is shown in Figure 3k. While many Penicillium spp. can cause post-harvest problems in cannabis inflorescences, there are a number of benefits that they confer as endophytes in numerous plants. Penicillium chrysogenum, reclassified as P. rubens [94], is known to produce the antibiotic penicillin [94,95,96] and has been recovered from the crowns and stem, petiole, and leaf tissues of cannabis plants [17,19,31]. In other crops, P. chrysogenum has been reported to confer various benefits, including biological control of plant pathogens [97,98,99] and the enhancement of plant development [100,101,102]. Further studies to evaluate other positive attributes of P. chrysogenum on cannabis plants are warranted. The second most prevalent species was P. citrinum, which has also been reported to be present in cannabis stem and petiole tissues [17,19] and in cannabis inflorescences [32]. This species is reported to function as a beneficial endophyte as well, providing plant growth promotion and biological control properties [103,104,105]. Although both Penicillium species may confer benefits as endophytes, they have also been recovered as epiphytes contaminating cannabis inflorescences from both indoor and outdoor crops [81]. The production of antibiotic compounds, including penicillin and citrinin, by these species may raise concerns if they are perceived to affect the health of cannabis consumers and are produced in sufficient quantities to be of concern [14]. They would therefore need to meet the regulatory guidelines outlined previously in order to be used on cannabis plants.

A second interesting microbe detected in all tissues, including seeds, was the mycorrhizal fungus R. irregularis, whose origin may be the growing substrate or seeds and cuttings as the substrate was not applied to the plants utilized in this study. This species is known to promote plant development in several ways [106,107], and it can confer bio-stimulant properties on cannabis plants [28,61]. Mycorrhizal-based inoculants were reported to enhance phytocannabinoid production when applied to the roots of cannabis plants [28,61], as well as promote plant growth [60,61]. The role of this mycorrhizal species in cannabis plants awaits further investigation, including the type of propagule (spores and mycelium) that may be present inside stems and leaves. Future research on endophytes to enhance cannabis growth should prioritize these two fungi because of their prevalence in cannabis tissues and their demonstrated positive effects from the published literature on other plant species. They should also be able to fulfill regulatory requirements more readily.

The common genera of yeasts that were present in cannabis tissues are shown in Figure 4. The majority were present in the growing substrate from which they are presumed to have originated, but others were likely introduced by air-borne sources. These included Aureobasidium, Candida, and Tilletiopsis. In field-grown hemp inflorescences, T. washingtonensis was found to be commonly present [12]. Yeasts, in general, do not pose concerns in cannabis quality assessments unless the populations exceed the total yeast and mold levels set by regulatory agencies [32], or the presence of some genera such as Candida poses a health concern for susceptible individuals. There are no previous reports utilizing yeast species for biological control or bio-stimulant research on cannabis or hemp.

The microbiome data presented in this study shed light on the widespread nature of endophytic fungal and yeast species shown to be present in greenhouse-cultivated cannabis tissues, some of which are potential plant pathogens, while others may be saprophytic or incidental in nature. The growing substrate had a significant impact on the types and quantity of microbes in these tissues. Plants initiated from cuttings taken from mother plants containing endophytic species would likely also harbor a similar background of fungal and bacterial species. With the exception of Beauveria, Mycosphaerella, and Sordaria, which were present in mother plants at low levels and not detected in cuttings derived from them, the carry-over population of microbes was significant. Similar to the high levels of transmission of important plant pathogens, such as F. oxysporum [81] and hop latent viroid [108] in cuttings taken from diseased mother plants, cannabis cuttings appear to also harbor a significant proportion of endophytes that are derived from the mother plants. Avenues to remedy this background microbiome are discussed below.