## Problem box Estimating Ne from information about N

Imagine that a conservation biologist approaches you asking for assistance in estimating the genetic impacts of a recent event in a captive population of animals housed in a zoo. The zoo building where the animals were kept experienced a fire, killing some animals outright and requiring the survivors to be relocated to a new enclosure that is not ideal for breeding. Before the fire, the population was stable at 30 males and 30 females for many generations with an effective population size of 60. After the fire there were 15 females and 10 males. Due to the disruption and relocation of the animals, breeding behavior changed. Before the fire variation in family size was Poisson distributed with a mean of 2.0. In the one generation after the fire, family size has a mean of 4.0 and variance of 6.5. What are the genetic impacts of the fire on the effective population size? What are some of the assumptions specific to this case used in your estimate of Ne?

explain Buri's results. Although there was a census size of 16 flies in each bottle, an unequal breeding sex ratio in each bottle could explain the higher rate of fixation and loss. For example, a breeding sex ratio of eight females and six males due to failure of some males to mate successfully each generation would give an effective population size of Ne ~ 14 using equation 3.44. It is also possible that there was a relatively high degree of variation in reproduction among the females. For example, if variance in family size was 3.5 and was combined with the effects of the unequal breeding sex ratio (using 14 instead of 16 for N(_i), equation 3.46 estimates that Ne ~ 10. It might also be that in a few of the generations the population size was smaller than intended due to mistakes when handling and transferring flies to new bottles. However, equation 3.42 shows that the effect of a population size of 14 for one generation out of19 is slight (Ne = 15.88). Therefore, infrequent fluctuating population sizes would probably have had only a minor impact on the results. Thus, the difference between the rate of fixation and loss in the Markov chain model and in the actual fly populations can be explained by several plausible factors that distinguish the census and effective population sizes.

## Post a comment