Table 3.
Model parameter estimates from Bayesian GLMMs relating measures of DWV diversity (Richness, Evenness, and the Shannon-Wiener index) explained by DWV log-loads and the degree of dominance of the single most dominant ASV in each sample (explanatory variables were mean-centred). Richness and Shannon-Wiener parameters are at the log-scale because a log-link was used in the modelling (Appendix 1). Estimates are presented with the means and standard deviations of the Bayesian posterior distributions as well as the Probability that their effect is in the direction indicated.
| Response variable | Estimates | |
|---|---|---|
| Richness | Parameter | Probability |
| Intercept | 2.180 ± 0.109 | – |
| DWV load | 0.033 ± 0.009 | 99.9% (positive) |
| Proportion dominant ASV | −0.018 ± 0.001 | 100% (negative) |
| Evenness | – | – |
| Intercept | 0.238 ± 0.007 | – |
| DWV load | 0.0006 ± 0.001 | 65% (positive) |
| Proportion dominant ASV | −0.009 ± 0.0003 | 100% (negative) |
| Shannon-Wiener index | – | – |
| Intercept | −0.825 ± 0.076 | – |
| DWV load | 0.018 ± 0.009 | 98.5% (positive) |
| Proportion dominant ASV | −0.034 ± 0.002 | 100% (negative) |