Table 5.
Statistical analysis of transmission experiments using CO2-treated leaves as virus source
| Experiment | n | Transmission frequency | LCI | UCI |
| Global | ||||
| Control | 16 | 0.13 | 0.09 | 0.17 |
| CO2 | 16 | 0.22 | 0.18 | 0.27 |
| Experiment 1 | ||||
| Control | 4 | 0.11 | 0.05 | 0.19 |
| CO2 | 4 | 0.15 | 0.07 | 0.25 |
| Experiment 2 | ||||
| Control | 6 | 0.14 | 0.08 | 0.21 |
| CO2 | 6 | 0.21 | 0.14 | 0.28 |
| Experiment 3 | ||||
| Control | 6 | 0.13 | 0.07 | 0.20 |
| CO2 | 6 | 0.30 | 0.22 | 0.39 |
We measured the transmission rate by aphids for each treatment condition (control or CO2) for the three experiments. As we found a non-significant interaction between experiments and treatments (GLM, df = 2, χ2 = 2.92, p=0.23), the three experiments were pooled for analysis (line named ‘global’). CO2 induced a significant increase in the transmission rate compared to the controls without CO2 (hierarchical GLM model, df = 1, χ2 = 9.12, p=0.0025), with 13.0% (95% CI: 9.4–17.0%) for control in ambient atmosphere and 22.4% (17.8–27.2%) for CO2-treatments.
n: number of repetitions per experiment; LCI, UCI: lower and upper limits of confidence intervals, respectively.