A population bottleneck is similar in important ways to the founder effect. Population bottlenecks occur when some environmental change leads to the dramatic reduction of the size of a population. Catastrophic environmental changes, such as asteroid impacts, massive and prolonged volcanic eruptions (associated with continental drift), or the introduction of a particularly deadly pathogen, which kills a high percentage of the organisms that it infects, can all create population bottleneck effects. Who survives the bottleneck can be random, due only to luck, or based on genetic factors (for example, leading to disease resistance).
There is compelling evidence that such drastic environmental events are responsible for population bottlenecks so severe that they led to mass extinctions. The most catastrophic of these extinction events was the Permian extinction that occurred ~251 million years ago, during which it appears that ~95% of all marine species and ~75% of land species became extinct.88 If most species were affected, we would not be surprised if the surviving populations also experienced serious bottlenecks. The subsequent diversification of the surviving organisms, such as the Dinosauria (which includes the extinct dinosaurs and modern birds) and the Cynodontia, which includes the ancestors of modern mammals, including us, could be due in part to these bottleneck-associated effects, for example, through the removal of competing species or predators. The Cretaceous-Tertiary event, which occurred ~65 million years ago, contributed to the extinction of the dinosaurs and led to the diversification of mammals (which had first appeared in the fossil record ~160 million years ago), particularly the placental mammals.
While surviving an asteroid impact (or other dramatic changes in climate) may be random, in other cases who survives a bottleneck is not. Consider the effects of a severe drought or highly virulent bacterial or viral infection; the organisms that survive may have specific phenotypes (and associated genotypes) that significantly influence their chance of survival. In such a case, the effect of the bottleneck event would produce non-random changes in the distribution of genotypes (and alleles) in the post-bottleneck population – these selective effects could continue to influence the population in various ways. For example, a trait associated with pathogen resistance may also have negative phenotypic effects. After the pathogen associated bottleneck, mutations that mitigate the resistance trait's negative effects (and may have their own effects) would be selected. The end result is that traits that would not be selected in the absence of the pathogen, are selected. In addition, the very occurrence of a rapid and extreme reduction in population size has its own effects. For example, it would be expected to increase the effects of genetic drift (see below) and could make finding a mate more difficult.
We can identify extreme population reduction events, such as founder effects and bottlenecks, by looking at the variation in genotypes, particularly in genotypic changes not expected to influence phenotypes, mating preference, or reproductive success. These so-called neutral polymorphisms are expected to accumulate in the non-sense (intragenic) parts of the genome at a constant rate over time (can you explain why?) The rate of the accumulation of neutral polymorphisms serves as a type of population-based biological clock. Its rate can be estimated, at least roughly, by comparing the genotypes of individuals of different populations whose time of separation can be accurately estimated (assuming of course that there has been no migrations between the populations). Such studies indicate that the size of the human population dropped to a few thousands individuals between ~20,000 to 40,000 years ago. This is a small number of people, likely to have been spread over a large area.89 This bottleneck occurred around the time of the major migration of people out of Africa into Europe and Asia. Comparing genotypes, that is, neutral polymorphisms, between isolated populations enables us to to estimate that aboriginal Australians reached Australia ~50,000 years ago, well before other human migrations90 and that humans arrived in the Americas in multiple waves beginning ~15,000 to 16,000 years ago.91 The arrival of humans into a new environment has been linked to the extinction of a group of mammals known as the megafauna in those environments.92 The presence of humans changed the environmental pressures on these organisms around the world.