Imagine a jar holding twenty marbles. Ten are red and ten are blue, representing two versions of a single gene within a population of organisms. Each generation requires picking one marble at random to create an offspring with the same color. This process repeats until the new jar contains twenty marbles again. Unless the second jar happens to hold exactly ten red and ten blue marbles, the ratio has shifted purely by chance. No organism was better or worse than another. The change occurred because random sampling altered the frequency of existing alleles. When few copies of an allele exist, this effect becomes more noticeable. Large populations experience less fluctuation due to the law of large numbers. Small populations can lose an entire variant in just a few generations if no individuals carry it into the next batch.
Mathematical Models And Simulations
Scientists use specific frameworks to calculate these probabilities. The Wright-Fisher model assumes non-overlapping generations where each copy of a gene is drawn independently from the previous pool. Ronald Fisher and Sewall Wright developed this approach to describe how allele frequencies shift over time. Another framework called the Moran model allows for overlapping generations. In the Moran model, one individual reproduces while another dies during each time step. This creates a tridiagonal transition matrix that simplifies mathematical solutions compared to the Wright-Fisher approach. Computer simulations often favor the Wright-Fisher model because fewer time steps are needed to complete a generation. Genetic drift runs twice as fast in the Moran model despite producing qualitatively similar results. These models help researchers predict when an allele will become fixed or lost within finite populations.