Stabilizing Selection on English Sparrows

During a severe winter storm in New England in 1898, 136 English Sparrows (Passer domesticus) succumbed to cold shock and literally fell out of trees. HC Bumpus collected the birds and took them to his lab, where 72 recovered and 64 died. Bumpus made a series of measurements on both groups [left]. His data tend to demonstrate that the standard deviation of each measurement (grey bars on either side of the mean) and hence the variance was greater in the birds that died, which suggested truncation of either tail of their distributions. This is summarized in the bar graph [right] for body size of females (measured as Length and Weight). Stabilizing Selection therefore operated to maintain population variation closer to the mean. Bumpus published his raw data measurements, and the "Bumpus data set" has been subjected to many subsequent re-analyses.