Table C.2.
Expected impact of assumptions on the ‘AI dynamics’ compartment of the model on the model outcome
| Assumption | Expected impact | Explanation |
|---|---|---|
| The entrance of migratory susceptible wild birds into the EU during migration season is modelled with a constant rate | Negligible impact |
1) The composition of the migratory wild birds does not affect the output 2) Different species arriving at different time covering all the migration period |
| Once birds land in the virtual area they stay there for the whole season (125 days duration) | Negligible impact | Birds will stay long enough in the area to allow transfer of the virus. Prolonging their stay has limited impact |
| Entry of infected migratory WB in the EU is clustered on one day | Assessed quantitatively | Worst‐case scenario |
| The distribution of birds by day of infectivity is uniform at the time they cross the EU border | Most realistic (negligible impact) | No evidence for a link between infection and host fitness |
| Once a wild bird got infected, it becomes able to transmit infection to a susceptible bird via its excretions from the next day | Most realistic (negligible impact) | Based on experimental data (Keawcharoen et al., 2008) |
| Random pattern assumed for the release of the infected excretions and for the contact birds‐excretions | Most realistic (negligible impact) | From data defecation rate, every 5 min, and they move randomly (e.g. about 7 min in Durant et al. 2006) |
| Contact with excretions is the only route of infection. | Most realistic (negligible impact) | Direct and indirect transmission of virus occurs via excretions |
| Groups of wild bird species are assumed to forage in a similar surface of the virtual contact area during a given day | Most realistic (negligible impact) | The probability of a wild bird to get in contact with contaminated excretions was estimated for groups of wild bird species with different behaviours |
| Probability of getting in contact with infected excretions excreted by birds from the same category is higher | Most realistic (negligible impact) | Birds of the same category share the same ecological niches which has only partial overlap with the other category |
| Quantity of excretion is not considered to differ between the different categories of birds | Negligible impact | Sharing the same ecological niches for birds of the same category has a much higher influence on the probability of contact compared to the quantity excreted |
| Infectivity of the excretions in the environment is assumed to be maintained for a number of days after release. During the persistence period level of infectivity of the excretions is assumed to be constant. | Worst‐case scenario | Evidence indicate a decrease of infectivity (Nazir et al., 2010) |
| Number of infected excretions is assumed to be directly proportional to the number of birds shedding the virus (infected) plus number of birds that recovered from infection but whose excretions are still infectious | Negligible impact | Sharing the same ecological niches for birds of the same category has a much higher influence on the probability of contact compared to the quantity excreted |
| AI infection confers immunity to birds once birds recover from the disease | Negligible impact | Quantity of virus excreted by a bird re‐infected with the same clade in a migration season is considered negligible to influence the outcome of the model; low prevalence low probability of reinfection |
| All migratory non‐water birds are susceptible when they enter the EU | Most realistic (negligible impact) | Categories were set up also on the probability they bring in the infection; evidence: very few non water birds found infected with HPAI 2.3.4.4 clade |
| All residential wild birds (both water birds and non‐water birds) are susceptible to the infection at the start of the migration season; the influence of heterologous immunity and previous exposure is not considered | Worst‐case scenario | Immunity would reduce viral amplification |
| Population dynamics of wild birds do not include baseline wild bird mortality | Worst‐case scenario | Including baseline mortality will decrease the size of the susceptible population. Analysis of the impact on the outcome of the model was negligible |
| Population dynamics of wild birds do not consider hatch | Most realistic (negligible impact) | Hatching does not occur during the migration or winter seasons |