Artificial Selection with Wisconsin Fast Plants
Can selection cause evolution in a single generation? Observing the effects of artificial selection in order to understand patterns of inheritance, simulating AP Biology Investigation #1
Background
Evolution by artificial selection is largely responsible for many of the most desirable traits in agricultural species of plants and animals that feed the modern world. Additionally, much of the diversity we see in pet species from dogs to pigeons is the result of human-directed evolution by artificial selection. When breeders and farmers select only individual animals or plants with desirable traits to reproduce, this can lead to the evolution of those species and the resulting and often dramatic change can occur rapidly when compared to evolution by natural selection. In this classic lab activity, students can observe artificial selection in action using Wisconsin Fast Plants. This lab is frequently done at AP Bio Lab Investigation #1, but could be appropriate in any middle or high school life science class. Additional complexity could be added for a university level course.
Students conducted this experiment, aligned with Investigation 1 in the AP Lab Manual, using Wisconsin Fast Plants (Brassica rappa) to test the extent to which artificial selection can cause observable evolution in a population within the span of a single generation. To conduct this test students first grew 205 plants in a parental generation up to day 15 and counted the number of trichomes (tiny hairs along the petiole or the margin of the first true leaf). Trichomes aid in plant defense against predators and are known to reduce insect feeding on plant tissues. These parental plants were then grown to seed. The seeds were only collected from plants with a trichome count of 25 or greater (~25% of the population).
Students then raised the seeds from the individual plants with the most trichomes to become an F1 generation (biology parlance for “first generation”) and grew them until day 15. At this time the number of trichomes were counted for 201 plants in the F1 generation.
Dataset
Variables
Generation: Either 1, for the first generation (F1) plants, or 2, for the second generation (F2) or offspring plants.
Date: The date on which the observation of Trichome Count was made.
Growth Day: The “age” of the plant; the observations of trichomes for the F1 and F2 generations were both made when the plants were 15 days old.
Plant #: The unique identity of each plant included in the study.
Trichome Count: Number of small hairs along the petiole of each plant; the trait being manipulated by artificial selection in the activity.
Parent: A categorical variable to indicate whether the plant being measured belonged to the parent generation or offspring generation.
Activity
Data aren’t always easy to interpret visually, especially since datasets with a large sample size (n) are better for making statistical inferences. Can you determine if there is a difference in mean trichome count just by looking at the data in the table view here? What is n for this activity?
Use the “Make a graph” feature to create a scatter plot of the data. Show Generation on the X-axis and Trichome Count on the Y-axis appropriately. Now what can you tell about the means between the two generations? Add descriptive stats with the checkbox to the right of the graph to highlight the mean for each group on the graph.
Let’s see if the difference in means is significant. Run a t-test to determine if the mean number of trichomes per individual between the parental and F1 generations is significantly different. Use the Graph Driven Hypothesis button to the right of the graph to do this. Is the difference significant?
Given an example of how the trichome count in Brassica rappa might hypothetically change through natural selection in a wild population rather than through artificial selection in this study population.