Measuring Selection in the Field

Readings - Introduction - Ideas for exercises

Below are some ideas for short exercises that can be used to demonstrate the effects of selection in the field.  Complete  outlines of exercises are not included, but instructors can shape these ideas to fit the needs of their students.


Manly, B.F.J. 1985. The Statistics of Natural Selection. Chapman and Hall: London.


Students often assume that the effects of natural selection are only measurable over large geologic time scales. This however, is not the case, and there are many ways in which a laboratory or field exercise can be designed to measure changes in populations of real organisms. In many cases, phenotypic variation in a natural population can be measured and the distribution of a particular characteristic can indicate how a population has been shaped by natural selection. For example, comparing a particular character of  living versus dead individuals in a given population can give an indication of the possible effects of predation on survivorship, etc. The following examples are intended to provide ideas for exercises in which students measure traits in natural populations and infer the effects of natural selection.

1 . Sampling from a population within a generation.   Sample individuals in a population from which birth, emigration and immigration are not possible. Take samples from two points in time of the same population. Differences in survival can be inferred to be the result of selection.

    a. Example (from Manly, 1985. p 47 ) "Predation of corixids by minnows". After Popham (1944) - Look at survival rates of corixid insects in a pond before and after the introduction of a predatory fish (minnows). Different characteristics of morphs of a single species, or differences between similar species could be measured prior to predator introduction and again after predator introduction.

    b. A similar sort of study could be easily accomplished with artificial prey and real predators. This sort of experiment would allow students to carefully control the characters that they are interested in. For example, a test of aposematic coloration in caterpillars or snakes could be done with birds. Students could create artificial pastry / play-dough caterpillars or snakes of varying colors and then place them in a natural environment to examine the effect of warning coloration on predation.

(see Pfennig DW, Harcombe WR, Pfennig KS, 2001. Frequency-dependent batesian mimicry - Predators avoid look-alikes of venomous snakes only when the real thing is around. Nature. 410: 323-323.)

2. Comparing live versus dead individuals in a population. In populations where dead and live individuals are easily collected, such as snails.

    a. A classic example is from thrush anvils, rocks where birds crush snail shells prior to eating. Differences in shell characteristics could be compared between the dead (predated) individuals found at the anvils and those live individuals collected away from anvils. Size distribution, color morph, etc could be compared. Size distribution could also be an indication of profitability of various sizes of snails as well as ease of opening and might be a good lab to investigate optimal foraging in real predators.

    b. Traditional mark-recapture techniques could also be employed. Large data sets of recovered bands could be obtained from the Bird Banding Laboratory (BBL) ( Waterfowl have the highest band recovery rate, and most of these bands are returned by hunters. Students could analyze trends in the size / sex of birds banded versus those recovered. Note: sufficient time to request data sets from the BBL should be allowed.

3. Non-random mating.

    a. Measurements of animals collected in copula (such as insects or, with some effort, frogs) can give indications of whether certain traits are predictive of higher reproductive (or at least mating) success. Also,  assortative mating can be investigated (including in humans  . . . though proxys for mating success need to be observable in public settings - hand holding, etc. Humans tend to mate assortatively according to height/size.)

    b. Measurements of flowers that are being visited by pollinators compared with those that are not. Some control of nectar volume depleted by previous foragers should be included  - covering several plants with mesh netting for a given amount of time to ensure that no flowers have been visited. Size, color, of flowers, relative flower location on a plant, could be measured.