Table 3. Feasibility of fixed control strategies adapted to the time and place where a simulated FMD epizootic begins.

	Control strategy
	SO	PS	PV	SPS	SPV	SPSV	SV
Direct costs
Initial conditions	1.4 (52%)	1.7 (59%)	0.8 (55%)	3.3 (58%)	0.8 (53%)	1.3 (58%)	1.2 (53%)
Residual variance	1.3 (42%)	1.2 (41%)	0.6 (45%)	2.4 (42%)	0.7 (47%)	0.9 (42%)	1.1 (47%)
Export losses
Initial conditions	0.4 (60%)	0.6 (64%)	0.3 (75%)	0.8 (60%)	0.3 (68%)	0.3 (68%)	0.3 (58%)
Residual variance	0.3 (40%)	0.3 (36%)	0.1 (25%)	0.5 (40%)	0.1 (32%)	0.2 (32%)	0.2 (42%)
Slaughtered herds
Initial conditions	1.1 (42%)	1.3 (57%)	0.6 (51%)	1.9 (55%)	0.6 (50%)	0.9 (55%)	1.0 (52%)
Residual variance	1.0 (48%)	1.0 (43%)	0.6 (49%)	1.6 (45%)	0.6 (50%)	0.8 (45%)	0.9 (48%)

Note: The table gives, for a given impact and for a specific control strategy, the parts of variance explained by the initial conditions (place and date where the epizootic begins) and by the intrinsic variability of epizootics (residual variance). Impacts correspond to three stakeholders: government (indicator: public costs), agro-food industry (indicator: export losses), and public opinion (indicator: number of slaughtered herds).SO: basic stamping-out strategy, PV: protective vaccination, PS: preemptive slaughter, SPV: selective protective vaccination, SPS: selective preemptive slaughter, SPSV: selective preemptive slaughter and preventive vaccination, SV: suppressive vaccination. Impact indicators have been log-transformed.