
use trait-distribution data to tell selection from drift, explain how a trait’s effect on survival changes its frequency, revise our model of the colony, and brainstorm options for the sanctuary.
Phase 3: Make Sense
Do Now. Five coins each land heads 7 times out of 10. Is that luck, or are the coins weighted? What would convince you?
Which pattern across replicate populations points to SELECTION rather than drift?
Selection or drift? Both change allele frequencies. Selection is consistent: populations in the same environment move the same way, because the trait affects survival and reproduction. Drift is random: populations move in different directions, and the change is largest where populations are small.
The dataset (sample data). Small mice live on pale sandy dunes and on dark inland soil. Coat color is inherited; owls and hawks hunt them. Biologists tracked the frequency of a light-coat allele in five dune populations of about 1,000 mice and in five small inland populations of about 30 mice.
| Population | Generation 0 | Generation 10 | Generation 20 |
|---|---|---|---|
| Dune 1–5 (mean, range) | 0.30 (0.28–0.32) | 0.44 (0.41–0.47) | 0.58 (0.54–0.62) |
| Inland 1 | 0.30 | 0.18 | 0.10 |
| Inland 2 | 0.30 | 0.41 | 0.45 |
| Inland 3 | 0.30 | 0.26 | 0.28 |
| Inland 4 | 0.30 | 0.12 | 0.00 |
| Inland 5 | 0.30 | 0.35 | 0.38 |
Worked example 1: describing the distributions. On the dunes, all five populations rose, and the spread stayed narrow (range 0.08 at generation 20). Inland, the five populations moved apart: the range grew from 0 to 0.45, one lost the allele entirely, and the mean (0.24) barely changed. The dune data show selection; the inland data look like drift.
Worked example 2: from survival to frequency. On the dunes, a trapping survey (sample data) found that over one season 60% of light mice and 45% of dark mice survived. Light mice were 1.33 times as likely to survive (0.60 ÷ 0.45). Because coat color is heritable, survivors pass on more light-coat alleles, so the trait distribution shifts toward light each generation. That is how a trait that affects survival changes a population, and why the dune mice became better matched to their background, an adaptation in anatomy (coat) shaped by selection.
Worked example 3: revising the colony model. Our 4.01 model said “the seals came back, so they are fine.” Version 2: Because only a few seals survived the hunting, the population lost many alleles in the bottleneck. Because a few dominant males father most pups, fewer individuals pass genes on each generation, so drift keeps working even in a large colony. Because group breeding has costs (crowded pups) and benefits (many females gathered where males defend them), behavior shapes both survival and the genetic future. And selection can act only on variation that exists, so low variation is a risk if a new disease or a changing ocean arrives.
Ideate. Now turn the science into options for the sanctuary. Rules: no idea is wrong in the first ten minutes; build on others’ ideas; aim for at least eight. Each idea must protect breeding behavior, track the colony’s numbers, or track its genetic health.
Use the dune and inland data.
Write the CER, then ideate.
Write a CER answering: did selection or drift change the dune and inland populations? Use at least three numbers, and explain how a trait's effect on survival shifts a trait distribution.
Write version 2 of your colony model using the frame “We used to think… The new evidence… Now we think…” Then list at least eight ideas for the sanctuary and star your team's top three.
We can tell selection from drift and we have a list of ideas. Next: we design competing plans for the sanctuary.
Exit ticket: name one pattern in data that would convince you a trait is being selected, and one that would suggest drift.
How confident are you that you can use trait-distribution data to explain whether selection or drift changed a population?
Materials: the dune and inland table, graph paper or spreadsheets, the 4.01 colony model posters, sticky notes for ideation.
Data note: the mouse data are sample data, loosely inspired by well-known studies of light-colored mice on pale sand dunes; no specific study’s numbers are used.
Listen for: treating one population’s loss of an allele as selection; “the mice turned light to hide”.
Support / stretch: give an organizer with columns trend, spread and verdict. Stretch: use 1 − 1/(2N) to estimate how much heterozygosity an inland population of 30 keeps over 20 generations (about 0.71).