*4.1. The Necessity of Characterizing Fungal Microbiomes in MOR and AOF Samples*

MOR and AOF are derived from the roots and fruits of plants, respectively. They are susceptible to fungal contamination during pre- and post-harvest processes owing to the lack of standard management. In our study, all 18 samples were contaminated with fungi. The relative abundances of fungi were varied in different samples. Ascomycota, Eurotiomycetes, Eurotiales, and Aspergillaceae were generally dominant at the phylum, class, order, and family levels in AOF and MOR samples, respectively. In general, *Penicillium* and *Cladosporium* were the dominant genera. In our previous studies, the predominant genus in four seed herbs (Platycladi Semen, Myristicae Semen, and Cassiae Semen) was *Aspergillus* [33–35]. The relationship between dominant fungi and herbs has not been well explained, owing to the complex factors that influence the fungal community. The storage conditions and matrix composition of CHMs may be responsible for the difference in fungal microbiomes between MOR and AOF. Moreover, the fungal microbiomes between moldy and non-moldy samples are different. All samples were clustered based on the presence or absence of visible molds in UPGMA, thus indicating the meaningfulness of grouping. *Cladosporium* and *Penicillium* were dominant in moldy and non-moldy samples, respectively.

Notably, all samples were contaminated with a potential toxigenic fungus, namely, *Trichoderma atroviride*. Additionally, four non-moldy samples were contaminated with another toxigenic fungus (*Fusarium equiseti*). Trichothecenes and zearalenone, which disturb hormonal balance and cause numerous diseases in the reproductive system, can be produced by *F. equiseti* [36,37]. The findings of our study are consistent with another study; potential mycotoxin-producing fungi were detected in non-moldy and moldy Ziziphi Spinosae Semen samples [38]. Similarly, Wei et al. detected AFs and OTA in six moldy and nine nonvisible moldy *Glycyrrhiza uralensis* samples collected from China [39]. Therefore, the safety assessment of CHMs should not depend on whether macroscopic molds are present or not. Characterizing fungal microbiomes, especially potential toxigenic fungi, in MOR and AOF samples is essential, which is providing an early warning for mycotoxin contamination.
