Zoology 510, Class Notes for Ridley, Chapter 11
The Analysis of Adaptation.

No written assignment, but you should try to answer the Study and Review Questions at the end of the chapter.

Brief Outline

• Overview of this unit (Part 3, Adaptation and Natural Selection, Chapters 11-13).
• Overview of Chapter 11. The Analysis of Adaptation.
• Checklist of important concepts and terms
• Section-by-section comments
  • 11.1. Adaptation may not be obvious.
  • 11.2. Methods of studying adaptation.
  • 11.3. The function of sex.
  • 11.4. Sexual selection.
  • 11.5. The sex ratio
  • 11.6. Different levels of understanding.

510 index page

Overview of Chapter 11.

• This chapter is one piece of a larger argument.  Before concentrating on details of this chapter, you should quickly review the entire unit (Chapters 11, 12, and 13).
• Issues presented in Chapter 12 (Units of Selection) and Chapter 13 (Adaptive Explanation) are quite relevant to understanding the example problems described in Chapter 11.
• The main point of Chapter 11 is, "The way organisms are adapted may not be obvious".  Everything else in the chapter is subservient to this one main point.  Your job is to understand how much a thorough understanding of adaptation actually entails.
• Three specific problems are used to illustrate the problem of adaptation.
  • Why is sexual reproduction so common?
  • Why do animals often display extreme sexual dimorphism in secondary sexual characteristics?
  • Why the is the sex ratio in animals almost universally 1:1?

• Often, these facts are treated simply as basic descriptive observations in biology.
• But each of these problems is, in its own right, a major conceptual challenge for evolutionary biology.
• After each of these examples (sexual reproduction, sexual selection, sex ratio), you should review what the basic problem is in terms of complete a adaptive explanation.
• Then you should realize that the same challenge, of how adaptation should be analyzed, applies to every case in which adaptation is proposed as an explanation.

CHECK LIST of important TERMS

• Adaptation
• Methods for studying adaptiveness of a trait
  • Postulate alternatives (genetic variants).
  • Develop hypothesis for adaptive function.
  • Test hypothesis.
    • Description (confirmation of prediction).
    • Experimentation.
    • Comparative method.
• Sexual / asexual reproduction
• 50% cost of sex.
• Recombination (elimination of mutation)
• Muller's ratchet (see notes below)
• Group selection (see Chapter 12 (Units of Selection))
• Parasite-host coevolution
• Red Queen hypothesis (see text p. 630)
• Sexual selection
  • Sexual dimorphism
  • Secondary sexual characteristics
  • Male competition
  • Female choice
  • Runaway selection
  • Handicap theory
  • Heritability of fitness
  • Parasite theory of sexual selection
• Sex ratio
  • Why ratio is typically 50:50
  • Deviations from 50:50 ratio

Chapter 11, Section-by-Section Comments

11.1. "The way organisms are adapted may not be obvious."

• This cuts two ways.  Sometimes apparently insignificant traits might actually be adaptations, while some traits that look like adaptations might not.
• Far too often in biology, "adaptation" is simply presumed.
• In the words of George C. Williams (in Adaptation and Natural Selection, 1966), "adaptation is a special and onerous concept that should be used only where it really necessary."
• To establish a trait as an adaptation, it is necessary to demonstrate several things:
  • Demonstrate that the trait is (or has been) variable.
  • Demonstrate that the putatively adaptive variant could actually contribute to fitness.
  • Demonstrate that the potential fitness contribution of the putative adaptation outweighs its cost.
  • Demonstrate that selection is actually operating (or has operated in the past) to favor the putative adaptation over other variants.
  • There are more, including identifying the appropriate unit of selection and understanding how selection for the adaptation could be undone by mutation or drift, but that's enough for now.

11.2. "Three main methods are used to study adaptation."

• Ridley begins with a list of three "conceptual stages" in the study of adaptation.  This is the basic scientific method, applied to adaptation.
  • Postulate alternatives (genetic variants).
  • Develop a hypothesis for adaptive function of one particular variant.
  • Test the hypothesis.
• Without such careful formulation and testing, hypotheses of adaptation are little more than wishful thinking and story-telling.
• Do not confuse these three stages (above) with the three methods (below) for testing hypotheses of adaptation.
  • Description (confirmation of prediction based on design criteria).
  • Experimentation (manipulation, with appropriate controls).
  • Comparative method (phylogenetic analysis and correlation).
• For each example in the following sections, try to discern what methods are being used.

• 11.3.1. "Sexual and asexual reproduction should be distinguished."
  This is the postulation of alternatives (stage one, section 11.2 above); one must consider how a population would fare if reproduction were NOT sexual.

• 11.3.2. "Sex has a 50% cost."
  • This is the fundamental difficulty.  Whatever advantage (if any) sex has, it must be sufficient to outweigh this huge cost, if sex is to be properly considered as an adaptation.
  • Even in hermaphroditic species, the cost is maintained by selection for balanced investment in male and female gamete production (see section 11.5, below)
  • What follows (11.3.3 to 11.3.6) explores the development of hypotheses (stage two, section 11.2 above) to explain what adaptive advantage might override this cost.

• 11.3.3. "Sex can accelerate the rate of evolution."
  Sex makes possible the recombination of alleles at various loci.  Bringing together favorable alleles, rather than waiting for coincidence mutation, can accelerate the pace of adaptation.  This does seem to be a definite and undisputed advantage for sex.
  • Unfortunately, sex can also break up favorable combinations of alleles.  So, this does not seem like a clear and universal benefit sufficient to outweigh the 50% cost of sex.

• 11.3.4. "Is sex maintained by group selection?"
  • The acceleration of evolution should be favored by group (i.e., species) selection, based on both a higher rate of speciation and lower rate of extinction.  These processes are the species-level analogues of reproduction and survival for individual selection.
    • One difficulty with this explanation is that individual selection is expected to be much faster and more effective than group selection.  Individual selection, therefore, should consistently override group selection whenever the two levels operate at cross purposes.
    • Some species (notably aphids) strike a "balance" between sexual and asexual reproduction.
      • Since both strategies are clearly available for individual selection, this is taken to imply that both strategies are advantageous for individuals.  (Otherwise, one or the other strategy would be eliminated by selection.)
      • Therefore, in aphids sexual reproduction must have a strong enough advantage for individuals to balance asexual reproduction.
      • If that individual advantage exists for aphids, a group selection explanation may not be needed elsewhere.
      • We still do not know what that individual advantage really is.
    • Arguments based on group selection cannot (yet) explain the function of sex (i.e., thay cannot explain how selection for accelerated evolution can override selection for efficient reproduction.)
    • But we still don't have a good individual-selection explanation, either.

     

• 11.3.5. "Sexual reproduction can enable females to reduce the number of deleterious mutations in their offspring."
  • This proposed function for sex is a complement to the one above.  Just as sex can bring together "beneficial" new mutations, it can keep together a set of established, successful alleles.
  • Otherwise, in asexual lineages, deleterious mutations could and would accumulate in spite of selection.
    • Occasionally, every offspring will carry at least one deleterious allele.
    • Under such circumstances, the best that selection can do is find the least disadvantaged genotype.
    • From that point on, the least-deleterious allele will be fixed.
    • As this process repeats, deleterious alleles will accumulate, until viability is compromised for the entire population.
    • This is sometimes called "Muller's ratchet" or "mutational meltdown" (mentioned by Ridley only under "further reading")
  • This hypothesis can be developed quantitatively.  At a high enough rate for deleterious mutations, sex can provide sufficient advantage to outweigh the 50% cost.  However, it is seems to require unrealistically high mutation rates (although mutation rates are notoriously difficult to measure).

• 11.3.6. "The coevolution of parasites and hosts may produce rapid environmental change."
  • If environments change fast enough, successful reproduction from one generation to the next may demand that offspring NOT resemble their parents too closely.
  • Under such circumstances, sex may be advantageous.
  • The difficulty is finding a change which is BOTH sufficiently extreme to override the 50% cost AND common enough to explain the near universality of sex.
  • Rapidly evolving parasites are proposed as a selection pressure upon the hosts which changes rapidly enough to fit the bill.  And parasitism (by viruses, bacteria, protists, etc.) is practically universal.
  • Detailed knowledge of the genetics of host-parasite interactions is still inadequate to confirm or refute this hypothesis.

• 11.3.7. "Conclusion: It is uncertain how sex is adaptive."
  • Note where this problem stands in relation to the conceptual stages of section 11.2, above.
  • Mostly, we have concentrated on hypothesis development.
  • With regard to hypothesis testing:
    • Description (testing by consistency with expectations) has not demonstrated an adaptive benefit sufficient to outweigh the cost.
    • Experimentation is not practical.  We lack the ability to manipulate sexuality as a reproductive strategy.
    • The comparative approach, which is necessarily indirect and circumstantial, provides evidence that sex must be advantageous (i.e., it is maintained under nearly all circumstances, while its loss appears to be a "terminal" adaptation").  Special cases like aphids provide suggestive evidence that an advantage to individuals exists, but do not reveal what that advantage is.
  • There remains a gnawing suspicion that, in the words of John Maynard Smith, "some crucial aspect of the problem has been overlooked."
    • Your professor suspects that what has been overlooked is the possibility of something like group selection (11.3.4) successfully suppressing the potential for asexual reproduction to evolve.
    • By this hypothesis, asexual reproduction would indeed evolve to take advantage of the 50% cost of sex, if only asexual variants appeared.  But sex has become is so enmeshed in other features of genetic architecture that the variants do not arise.
    • An analogy can be drawn with cancer.  Here, selection based on immediate advantage for individual cells can indeed be very rapid and can overide advantage to the organism.  But organisms have evolved mechanisms to suppress the appearance of independently evolving cells.  Short term advantage (cancer growth) can indeed win out, but only if it's allowed to get started.  By analogy, then, asexuality may also be a winning strategy (as implied by the 50% cost of sex) in the short term but terminal (leading to extinction) in the longer term (as implied by the limited phylogenetic distribution of asexual species); therefore, those lineages which persist are those which successfully suppress the origination of asexual variants.
    • This hypothesis requires some special pleading to dodge George Williams "balance" argument.

• 11.4.1. "Sexual characters are often apparently deleterious."
  • Section 11.4 discusses examples where adaptive explanation based on sexual selection for the character must explain how ordinary selection against some deleterious effect of the character can be overcome.
  • Sexual characters which are NOT deleterious pose no explanatory difficulty; these will not be discussed here even though there is a rich literature.
  • Note how both experiment and comparative study have been used to support the following hypotheses.

• 11.4.2. "Sexual selection acts by male competition and female choice."
  • Comparative study, beginning with Darwin, provides extensive evidence that sexual dimorphism is more pronounced in species where individuals compete more strongly for mates, and the the sex (usually male) which competes has the more extreme character development.

• 11.4.3. "Females may choose to pair with particular males."
  • When females are passive, and males simply fight with one another for access, explanation of adaptation by sexual selection is straightforward and relatively unproblematic.
  • However, when females actively choose, and males compete to be chosen, it becomes somewhat more difficult to understand why females would choose males with deleterious traits.
    • Once the mating pattern is established, then females become obliged to choose the "preferred" type of male, or their male offspring will be less successful at finding mates.  So, it is fairly easy to understand how female choice even for otherwise deleterious traits can be maintained in a population after it has become established.
    • The problem is understanding how selection could establish such a mating pattern in the first place.  Sections 11.4.3 to 11.4.7 present several models (hypotheses) for how this could work.

  • The first hypthesis is Fisher's model.
    • An early form of the trait in question (ancestrally, prior to the effect of sexual selection) would be advantageous if larger.
    • Female choice evolves to prefer this the more advantageous, larger trait, because offspring of such females are more fit.
    • Once such female choice has become instinctive, males then face selection pressure from that choice, to elaborate the trait.
    • The resulting instinctive preference for "larger" will not automatically stop once "large enough" has been reached.
    • The result is runaway sexual selection, in which pressure of female choice drives male trait development beyond the value which would give optimal population fitness.
    • Equilibrium will be reached when the cost of the excessive trait balances the female preference.  At this stage, the trait no longer has any intrinsic fitness value for the population.

• 11.4.4. "Females may prefer to pair with handicapped males, because the male's survival indicates his high quality."
  • The second hypothesis is Zahavi's "handicap" model.
  • If a male can function in spite of some extravagant handicap, other genes for fitness must be in pretty good shape.
  • The "handicap" then serves as a signal for overall fitness.  (Human display can serve an analogous role, signalling economic fitness.  Men with low incomes cannot buy fancy clothes and sports cars.  So women who marry flashy men may have some increased expectation of economic resources for child support.)
  • Females who prefer males with the "handicap" will assure that all of their offspring get good genes (except, of course, those for the handicap itself).
  • In this model, the selected trait is costly, for both sexes, from the beginning.  But females benefit in other ways from using the costly trait as a signal.

• 11.4.5. "Female choice in Fisher's and Zahavi's models must be open-ended, and this condition can be tested."
  • In Fisher's model, "runaway selection"depends on positive feedback, in which females consistently prefer a more extreme form of the selected trait (rather than some particular optimal value).
  • In Zahavi's model, females must prefer a larger over a smaller handicap in order to maintain the handicap against selection to reduce it.
  • In fact, experimental manipulation has repeatedly demonstrated that females may prefer trait forms far more extreme that those presented by actual males.  (This is unsurprising, given that evolved instincts simply need to work; they need not be intelligently farsighted).
  • Such experiments show that either Fisher's or Zahavi's model could be true, but do not distinguish between the two models.

• 11.4.6. "Fisher's theory requires heritable variation in the male character, and Zahavi's theory requires heritable variation in fitness."
  • In Fisher's model, without heritable variation in the male character there would be no payoff for females who were choosy (their sons would be no more likely to mate successfully than sons of less choosy mothers).  So the cost of the character should lead to its reduction.
  • In Zahavi's model, without heritable variation in fitness there would be no payoff for females who preferred the costly handicap.  So the cost of the handicap should lead to its reduction.
  • Heritability is difficult to measure, so these expectations have not been extensively tested.

• 11.4.7. "Females may choose to pair with healthy, unparasitized males."
  • This hypothesis, by Hamilton and Zuk is a special case of Zahavi's model, in which a conspicuous male trait provides a fairly direct signal of health rather than an indirect signal of "good genes".
  • Some limited experimental evidence supports this hypothesis.  Whether it will hold more generally remains to be tested

• 11.4.8. "Conclusion: the theory of sex differences is well worked out but incompletely tested."

• 11.5.1. "Natural selection usually favors a 50:50 sex ratio."
  • Under most circumstances, a 50:50 sex ratio (males:females) should be an equilibrium.
  • Any departure from this ratio should create conditions (negative feedback) which will restore the equilibrium.
  • A lower proportion of either sex will make that sex disproportionately responsible for the next generation, so selection should favor greater production of that sex, as long as competition for mates is distributed among a large population.
  • Not presented in the text:  A similar argument supports the expectation of equal investment in male and female gamete production in hermaphroditic species.
  • Note that "group selection" would be expected to favor a much reduced proportion of males, since the advantage of sex (whatever it is) could be retained while increasing the efficiency of reproduction.  That this has not happened implies that group selection must not be especially effective.

• 11.5.2. "Local mate competition causes deviation from a 50:50 sex ratio."
  • The 50:50 ratio is at equilibrium only if there is a population-wide competition for mates.
  • Under special circumstances, such as in some parasitic wasps where mate competition is not population wide but brothers compete primarily with one another to mate with their own sisters, selection will favor a reduced proportion of males.  The optimal proportion of males will depend on the probability that brothers will compete only with another or with one or more additional sets of siblings.
  • Observed ratios correspond well with theoretical predictions.

• 11.5.3. "Trivers and Willard identified another case in which sex ratio should deviate from 50:50."
  • In red deer, data suggest that sons of dominant females have higher fitness than daughters, because they outcompete other males for access to many females.
  • Under such circumstances, it should be advantageous for dominant females to produce a higher proportion of sons than daughters, and data confirm this expectation.

• Sex ratio adaptation seems to be understood in quantitative detail.
• Adaptation by sexual selection has a well-developed set of theoretical expectations which have not been thoroughly tested.
• The adaptive function of sex itself remains deeply problematic and controversial.

Item of possible interest. Expert, up-to-date discussion of current ideas and research related on several of the topics of this chapter can be found in a special issue of Science, 25 September 1998 (vol. 281, pp. 1979-2008), devoted to the evolution of sex.

510 index page

Comments and questions: dgking@siu.edu
Department of Zoology e-mail: zoology@zoology.siu.edu
Comments and questions related to web server: webmaster@science.siu.edu