Table 2 Samples were partitioned by sampling strategy (swabs, different tape depths) and the bacterial composition was analysed against inter-individual variation, sampling location, skin status and skin group individually using PERMANOVA. Similarly, differences in ASV richness associated with explanatory variables was analysed for each sampling strategy separately using mixed linear modelling

Swabs

F-model

2.187

1.111

R2

0.595

0.051

P-value

0.001

0.001

Tape-1

F-model

1.726

1.12

R2

0.552

0.053

P-value

0.001

0.28

Tape-5

F-model

1.284

1.095

R2

0.541

0.061

P-value

0.001

0.018

Tape-15

F-model

1.197

1.056

R2

0.531

0.06

P-value

0.001

0.345

Tape-35

F-model

1.147

1.025

R2

0.558

0.065

P-value

0.001

0.416

Swabs

AIC

10320270

1315.8

LRT

1070.3

22.93

P-value

0.000

0.000

Tape-1

AIC

2327519

1106.3

LRT

870.5

2.32

P-value

0.000

0.678

Tape-5

AIC

1231403

912.1

LRT

703.78

0.25

P-value

0.2733

0.8836

Tape-15

AIC

571365

846

LRT

640.51

2.77

P-value

0.023

0.597

Tape-35

AIC

472461

757.3

LRT

572.04

3.97

P-value

0.044

0.41