Lecture XVIII - Sexual selection continued

30 October 2001


Sexual selection (continued)

Last time - intrasexual selection (e.g., male-male competition)

This time - intersexual selection (e.g., mate choice)


Concepts under development

1) mate choice exists

2) this can set up genetic correlations (linkage disequilibrium)

what is linkage disequilibrium? simply put, it is the non-random association between a chromosomes genotype at one locus and its genotype at another locus. In other words, certain combinations of alleles are more likely to be found together on a chromosome than other combinations of alleles.

Example 1: Female choice in long-tail widow birds

Observation: Male widow birds have very long tails

Hypothesis: Males have long tails as a result of a history of female preference for long tails

Prediction: Males with longer tails will get more mates


1) artificially create birds with short tails (cut tails off)

2) artificially create birds with long tails (glue pieces from step 1 onto some subset of birds)

3) control 1 = no manipulation; control 2 = cut off tail and re-glue tail back on (determine cutting effect)

4) Determine mating success of each of the 4 groups of males before and after manipulation.


Figures reproduced from Anderson, M. (1982) Nature 299: 818-820

Conclusion: Mate choice exists. Females prefer to mate with males with longer tails.


Example 2: Female choice in humans

Observation: MHC (major histocompatibility complex) type affects mate choice in mice, apparently through its effects on body odor.

MHC plays an important role in immume recognition by binding to foreign proteins

Question: Does MHC type also affect mate choice in human females?

Prediction: Human female college students will prefer odors of male students with different MHC types

What do we really mean by preference?

How would you assess preference?

What kind of experiment would you design to clearly test this prediction?


Wederkind et al. (1995) Proc Roy Soc Lond., B (on reserve in Life Sciences Library) (Link to pdf copy)

1) 49 female students and 44 male students were typed for thier MHC complex.

2) males were asked not to shower for 2 days and wear a standard t-shirt while sleeping

3) women were asked to rate teh odor of 6 t-shirts. Wederkind recorded which women were taking oral contraceptives


a) MHC dissimilar t-shirts were preferred (like a boyfriend) by women not taking oral contraceptives

b) MHC similar t-shirts were preferred (like a relative) by women taking oral contraceptives

Why are the results dependent on contraceptive usage?


Runaway sexual selection

1) Let us assume for now that there is no cost to some male trait, say tail length in widow birds

2) Let us further assume that some females prefer males with this trait. Other females do not show a preference. The preference is for relatively high values.

3) Finally, let us assume a large population (little chance of inbreeding)

4) We ask, what are the consequences of having choosy females in the population?

a) there will be a linkage disequilibrium generated between the female preference and the male trait in the next generation (i.e., preference alleles more likely to be found with alleles for exaggerated male phenotype in the population)

b) preference alleles 'hitch-hike' thorugh a correlated response to selection for the advantageous male trait

c) generates 'runaway sexual selection' --> process reinforces itself every generation with prevalence of preference making the exaggerated male trait even more advantageous. The population enters a positive feedback loop.

d) Continues until the population runs into some natural selection barrier that limits the extremeness of the male trait (e.g., tail feahers too long to evade predators)

Why doesn't the preference become less favored as the trait becomes more extreme? Females that prefer another trait will suffer a relative fitness cost because their sons will possess an unattractive trait to the rest of the females

See box 9.3 and figure 9.22 for a discussion of stalk-eyed flies

5) In groups, we designed an experiment to test the prediction that the male trait and female preference become genetically correlated within individuals (i.e., males with extreme traits also have alleles for strong preference and females with strong preference also have alleles for the extreme trait).