Table 4. Results of laboratory experiment statistical analyses.
| Year | Initial disease status | Disease metric | Test | df | F | P |
|---|---|---|---|---|---|---|
| 2009 | uninfected | Total prevalence | diel-cycling hypoxia | 2,12 | 9.19 | 0.004 |
| MHprevalence | diel-cycling hypoxia | 2,12 | 3.72 | 0.055 | ||
| intensity | diel-cycling hypoxia | 2,11* | 0.15 | 0.277 | ||
| mortality | diel-cycling hypoxia | 2,12 | 0.08 | 0.927 | ||
| shell growth | diel-cycling hypoxia | 2,12 | 0.66 | 0.534 | ||
| 2009 | infected | Total prevalence | diel-cycling hypoxia | 2,12 | 0.05 | 0.954 |
| MHprevalence | diel-cycling hypoxia | 2,12 | 0.03 | 0.972 | ||
| intensity | diel-cycling hypoxia | 2,12 | 0.15 | 0.861 | ||
| mortality | diel-cycling hypoxia | 2,12 | 0.98 | 0.403 | ||
| shell growth | diel-cycling hypoxia | 2,12 | 0.53 | 0.600 | ||
| 2010 | uninfected (ANOVA) | Total prevalence | diel-cycling hypoxia | 2,12 | 0.03 | 0.720 |
| MHprevalence | diel-cycling hypoxia | 2,12 | 0.05 | 0.95 | ||
| intensity | diel-cycling hypoxia | 2,12 | 0.56 | 0.587 | ||
| mortality | diel-cycling hypoxia | 2,12 | 0.08 | 0.924 | ||
| shell growth | diel-cycling hypoxia | 2,12 | 13.81 | 0.001 | ||
| wet weight growth | diel-cycling hypoxia | 2,12 | 4.13 | 0.043 | ||
| uninfected (ANCOVA) | Total prevalence | Model | 7,7 | 5.20 | 0.02 | |
| Diel-cycling hypoxia | 2,7 | 5.97 | 0.031 | |||
| P. marinus density in paired tank | 1,7 | 15.24 | 0.006 | |||
| Room position (block) | 4,7 | 3.26 | 0.083 | |||
| infected | Total prevalence | diel-cycling hypoxia | 2,12 | 4.17 | 0.042 | |
| MHprevalence | diel-cycling hypoxia | 2,12 | 0.13 | 0.881 | ||
| intensity | diel-cycling hypoxia | 2,12 | 0.85 | 0.454 | ||
| mortality | diel-cycling hypoxia | 2,12 | 3.98 | 0.047 | ||
| shell growth | diel-cycling hypoxia | 2,12 | 1.42 | 0.280 | ||
| wet weight growth | diel-cycling hypoxia | 2,12 | 0.79 | 0.477 |
One-way ANOVA was used to test effects of diel-cycling hypoxia on total prevalence and MHprevalence of P. marinus infections, as well as oyster mortality in both years. Nested ANOVA was used to test for effects on infection intensity and oyster growth. In addition ANCOVA was used to test for effects of diel-cycling hypoxia on total prevalence and growth in 1yo oysters for the 2010 experiment. For 3yo oysters in 2010, significant DO effects do not provide evidence for a negative effect of diel-cycling hypoxia; growth in the 0.5 mg l-1 treatment did not differ from that in controls, and mortality was higher in controls than in the 0.5 mg l-1 treatment.
*Control replicate with zero prevalence was not included in intensity analysis, resulting in a lower df.