Tolerant versus sensitive genomes:: The impact of deleterious mutation on fitness and conservation

Using two alternative mutational models, I investigate equilibrium diffusion predictions for population fitness. In the classical "sensitive model", fitness is highly sensitive to mutations, these usually having mildly detrimental effects. In the "tolerant model", most mutations have only tiny deleterious effects, but a small proportion is considerably detrimental. When the same relationship between the degree of dominance and the homozygous deleterious effect of mutations is assumed, both models predict important inbreeding depression after bottlenecking in large populations, although the sensitive one accounts for a higher average degree of dominance. Under the sensitive model, the rate of fitness decline due to deleterious fixation is large for effective population sizes in the tens, and could be important in the long term for effective sizes about 100, in agreement with previously published results. This suggests that conservation programs should act to avoid mutational meltdown. Under the tolerant model, however, the fitness decline due to deleterious fixation is generally low, indicating that conservation programs should give priority to avoid quick inbreeding, loss of genetic variability and adaptation to captive conditions, even if this reduces the strength of selection against new mutations.