We then assume that the decision-maker can predict the (mean) extinction time (T_{mathrm{ex}}) of (mathrm{E} [ N_{t} ]): T_{mathrm{ex}} = inf bigl{ t |mathrm{E} [ N_{t} ] = 0, t ge 0 bigr}.
We used λ as it covaried with other population viability indices, such as the exponential rate of increase, the median extinction time and the mean extinction time (0.53<r<1.00, all p<0.05).
This temporal autocorrelation in the impact of poisoning, the most realistic situation (scenario 2 Table 2), increased the extinction probabilities of the whole population to its maximum value (pe(t = 50) = 1, mean extinction time: 10.2 yearsee.28, scenarioario 2 Fig. 5).
This argument is based on the increase of the mean extinction time relative to unlinked selection.
A final statement about the relative importance of invasion probability vs. mean extinction time is not appropriate at this point.
This would require extensive numerical work, along with a derivation of a diffusion approximation to the mean extinction time for tight linkage.
In contrast, we have shown that the relative advantage of linkage in terms of the mean extinction time is maximized at an intermediate migration rate.
Because, for large populations, absolute extinction times become very large as the migration rate decreases, the biological relevance of this comparison may be confined to cases in which the mean extinction time of an unlinked mutation is not extremely high.
However, for small populations, our results show that linked selection can increase the mean extinction time to an extent that is biologically relevant, while, at the same time, not affecting the invasion probability much.
