Table 3. Likelihood ratio tests for honey residue responses to oilseed rape crop cover.
| Model | Log Lik. | Para. | φ | p | n | AICc | LRT |
|---|---|---|---|---|---|---|---|
| μ = a0 + a1*OSR | -64.87 | 4 | 0.92 | 1.12 | 109 | 139.6 | χ21 = 12.4, p<0.001 |
| μ = a0 + a1*Winter cereals | -69.02 | 4 | 1.03 | 1.14 | 109 | 146.4 | χ21 = 4.08, p = 0.04 |
| μ = a0 + a1*Arable | -67.73 | 4 | 1.01 | 1.13 | 109 | 143.8 | χ21 = 6.66, p = 0.01 |
The significance of the response of combined neonicotinoid residues detected in honey (NNI = clothianidin + thiamethoxam + imidacloprid) in the first year (2015) of the European Union moratorium on the use of neonicotinoids. Combined neonicotinoids residues are correlated with potential agricultural sources of these pesticides in the form of oilseed rape cover (OSR), winter sown cereal cover and total arable cover. Likelihood Ratio Tests assess whether these responses explain more variance than intercept only model. Log Likelihood, number of parameters (including φ and p for the Tweedie distribution), sample size (n) and AICc are also provided.