Table 1.
Distance-based linear modelling relating physical and biological predictors (appendix S3) of kelp forest responses to the heatwave.
| Marginal tests | ||||
|---|---|---|---|---|
| Predictor variable | SS(trace) | Pseudo-F | P | Variation explained |
| Genetic diversity (He) | 13334 | 58.00 | 0.0002 | 85.3% |
| Heatwave | 4732 | 4.34 | 0.07 | 30.2% |
| Nutrient concentration | 4059 | 3.51 | 0.09 | 26.0% |
| Reef topography | 1607 | 1.15 | 0.31 | 10.3% |
| Turf/foliose seaweeds | 1090 | 0.75 | 0.40 | 7.0% |
| Wave exposure | 788 | 0.53 | 0.48 | 5.0% |
| Fish herbivores | 165 | 0.11 | 0.76 | 1.1% |
| Depth | 43 | 0.03 | 0.87 | 0.3% |
| Best model selection | ||||
| AICc | r 2 | Predictor variables | ||
| 68.4 | 0.85 | Genetic diversity | ||
| 69.6 | 0.88 | Genetic diversity, Reef topography | ||
| 70.3 | 0.87 | Genetic diversity, Heatwave | ||
| 70.9 | 0.87 | Genetic diversity, Nutrient concentration | ||
| 71.0 | 0.87 | Genetic diversity, Turf/foliose seaweeds | ||
Top half: Marginal tests ascertaining the relationships to individual predictors (total trace = 15633). Bottom half: Multiple regression to ascertain the best (lowest AICc) combinations of predictors, showing the five best models overall.