Wood et al. 10.1073/pnas.0700062104. |
Table 2.
Results of an ANOVA after backwards elimination in which predictors explain 17% of the variance in the response, mass of macroalgae consumed by 55 L. littorea individuals maintained in the laboratory
Source | df | Mean square | F | P |
Whole model | 4 | 0.208 | 2.602 | 0.047 |
Block | 2 | 0.205 | 2.561 | 0.087 |
Shell length | 1 | 0.326 | 4.074 | 0.049 |
Infection status | 1 | 0.339 | 4.238 | 0.045 |
Error df for all sources = 50.
Table 3. Results of an ANOVA after backwards elimination in which predictors explain 77% of the variance in the change in percent cover of ephemeral macroalgae in experimental cages in the field
Source | df | Mean square | F | P |
Whole model | 14 | 0.025 | 2.696 | 0.053 |
Treatment | 1 | 0.083 | 9.057 | 0.012 |
Shell length | 1 | 0.041 | 4.425 | 0.059 |
Block | 6 | 0.039 | 4.264 | 0.018 |
Block ´ treatment | 6 | 0.026 | 2.867 | 0.062 |
Analysis compares infected versus uninfected treatments. Error df for all sources = 11.
Table 4. Results of an ANOVA after backwards elimination in which predictors explain 56% of the variance in the change in the first principal component of macroalgal community composition
Source | df | Mean square | F | P |
Whole model | 7 | 2.682 | 3.234 | 0.021 |
Treatment | 1 | 4.397 | 5.301 | 0.034 |
Block | 6 | 2.586 | 3.118 | 0.028 |
Analysis compares infected versus uninfected treatments. Error df for all sources = 18.