Table 2. Pairwise dissimilarities calculated among replicates (i.e., pseudo- β diversity) to evaluate the effect of random sampling and technical artefacts.

Pairwise dissimilarity	Description of calculation	Interpretation of dissimilarities
Within indexed primers	Between communities obtained using identical primer indices but sequenced in different MiSeq run (i.e., primer1/run1 vs. primer1/run2; primer2/run1 vs. primer2/run2)	Result from random sampling during sequencing only
Between indexed primers	Between communities obtained using different primer indices regardless of the sequencing run (i.e., primer1/run1 vs. primer2/run1; primer2/run1 vs. primer1/run2)	Result from a combination of technical artefacts during PCR (e.g., amplification bias caused by primer indices) and random sampling during sequencing
Within Illumina MiSeq run	Between communities obtained within the same MiSeq run regardless of the indexed primer set used (i.e., primer1/run1 vs. primer2/run1; primer2/run1 vs. primer3/run1)
Between Illumina MiSeq run	Between communities obtained in different MiSeq runs regardless of the indexed primer set used (i.e., primer1/run1 vs. primer1/run2; primer1/run1 vs. primer2/run2)