Table 4 The best-fit model for Shannon’s diversity index score, which explained 27% of within-sample diversity variance and had a p value of 1.04 × 10⁻⁰⁹. Slopes are given for continuous predictor variables and interactions between continuous and categorical predictors with two levels. Observed trends, from low to high average Shannon’s diversity scores, are given for the categorical predictors

1

0.089

0.089

17.62

< 0.001

0.0006

1

0.041

0.041

8.20

0.005

Surface > air

3

0.102

0.034

6.75

< 0.001

Winter > autumn > summer > spring

1

0.029

0.029

5.74

0.017

− 0.0017

1

0.005

0.005

1.04

0.309

− 0.1415

1

0.002

0.002

0.47

0.493

− 0.1018

1

0.002

0.002

0.43

0.514

− 0.0001

1

0.000

0.000

0.000

1.000

Outdoor > indoor

1

0.097

0.097

19.11

< 0.001

− 0.0130

1

0.051

0.051

10.02

0.002

− 0.0047

1

0.029

0.029

5.75

0.017

0.0005

3

0.061

0.020

4.04

0.008

3

0.034

0.011

2.24

0.085

3

0.029

0.010

1.91

0.130

1

0.005

0.005

1.01

0.316

0.0011

3

0.006

0.002

0.37

0.773

1

0.000

0.000

0.08

0.772

0.0060

1

0.000

0.000

0.00

0.990

0.0004

209

1.056

0.005