Table 1. The influence of risk factors on the calculated environmental impact quotient (EIQ) as determined by simulation analysis (N = 100,000).
| Median EIQ value | |||||
|---|---|---|---|---|---|
| Risk factor a | Low risk (1) | Medium risk (3) | High risk (5) | Relative influence b | Explained variance c |
| Dermal toxicity (DT) | 34.7 | 48.7 | 63.3 | 7.0 | 0.217 |
| Chronic toxicity (C) | 32.7 | 48.7 | 65.0 | 7.8 | 0.275 |
| Fish toxicity (F) | 45.3 | 47.3 | 49.3 | 1.0 | 0.005 |
| Leaching potential (L) | 46.7 | 47.3 | 48.0 | 0.4 | 0.001 |
| Surface loss potential (R) | 45.3 | 47.3 | 49.3 | 1.1 | 0.006 |
| Bird toxicity (D) | 42.7 | 48.0 | 52.7 | 2.5 | 0.028 |
| Soil half-life (S) | 44.0 | 48.0 | 51.3 | 2.0 | 0.018 |
| Bee toxicity (Z) | 43.3 | 48.0 | 51.3 | 2.0 | 0.018 |
| Beneficial arthropod toxicity (B) | 40.7 | 47.3 | 54.7 | 3.3 | 0.049 |
| Plant surface half-life (P) | 34.0 | 60.0 | na d | 13.0 | 0.283 |
aThe EIQ also includes “systemicity” (SY), but this risk factor is set to 1 for all herbicides; it was therefore not included in the simulation.
bRelative influence is the expected change in EIQ as the risk factor value increases by 1 unit, if all other risk factors are allowed to vary independently. Multiply the relative influence value by 2 to estimate the EIQ change per increase in risk factor category (low to medium, or medium to high).
cThe proportion of variation in the EIQ explained by each risk factor was determined by dividing the sums of squares for each risk factor by the total sums of squares for an additive model. Residual sums of squares = 0.100.
dNot applicable; the high risk category for plant surface half-life is not used in the EIQ calculation for herbicides.