Pest introductions via trade in tree seed may result from a lack of adequate survey and validation protocols. Developing better diagnostic protocols to identify potentially harmful pests and pathogens in forest tree seed is of critical importance. High-throughput sequencing-based barcoding and metabarcoding provide effective tools for screening potentially harmful organisms in various plant materials, including seeds. However, the sample size needed to detect the total microorganism diversity of a community is a major challenge in microbiome studies. In this work, we examined how increasing sample size (ranging between 100 and 1000 seeds) influences diversity of fungal communities detected by high throughput sequencing in Pinus sylvestris seeds. Our results showed that as sample size increased, fungal alpha diversity also increased. Beta-diversity estimators detected significant differences between the mycobiota from different samples. However, taxonomic and functional diversity were not correlated with sample size. In addition, we found that increasing the number of PCR replicates resulted in a higher abundance of plant pathogens. We concluded that for the purpose of screening for potentially harmful pathogens using HTS, greater efforts should be made to increase the sample size and replicates when testing tree seed.