ZOOL 304
Question and Answer
Ask a question, either in class or via e-mail. If the question appears to be of general interest, an attempt at an answer will be posted here.
Click on a question to see the response. As time passses, new questions will be added to the top of the list.
3/26/03 How can evolution explain homosexuality?
3/05/03 What's the point of separate sexes (rather than hermaphrodites)?
2/10/03 What should I expect on the exam?
2/10/03 How does "Hardy-Weinberg" relate to drift?
2/10/03 How does genetic drift relate to variation, change, and mutation?
2/10/03 In genetic drift, how are probabilities and allele frequencies related?
2/10/03 What do introns, exons, pseudogenes, etc. have to do with evolution?
2/10/03 Are all genetic traits correlated with reproductive success?
2/10/03 Which type of selection is responsible for most adaptive features?
2/10/03 What is coevolution?
2/10/03 What causes genetic change?
304 index pageHow might evolution explain homosexuality?
This question arises from the obvious presumption that homosexuality ought to reduce fitness. Why should a trait occur at a fairly high frequency (I've seen estimates as high as 10% of the general population), when bearers of that trait reproduce at a lower rate (or not all)? There are several plausible hypotheses (each with variations) that may be offered.
Caveat emptor. Any explanation for any human behavior is fraught with hazard, since such explanations are generally difficult (or impossible) to test experimentally and since they can be wielded for or against particular political / moral / religious / philosophical positions. This is most especially true for human sexual behavior, where "scientific" explanations can be used to justify the "naturalness" or "unnaturalness" of particular behaviors (and, by commonly held implication, the respective "normalcy" or "perversion" of those behaviors).
First, the presumption that homosexuality reduces fitness might simply be false. Perhaps homosexuality carries no significant fitness cost. Bonobos famously indulge in homosexual activity, but this seems not to be an ineffective or "abnormal" reproductive behavior but rather something more like play and social reinforcement. In other words, finding sex enjoyable has rather obvious fitness advantages, and homosexual behavior might simply be a "neutral" side effect of finding sex enjoyable. If so, then no further evolutionary explanation would be necessary. Furthermore, among people through much of "civilized" human history, women have not been free to choose their sexual partners and men may have been socially constrained to behave as heterosexual males (i.e., by fathering children) regardless of their sexual preferences. Thus we have another reason to suspect that fitness for individuals with homosexual tendencies might be just about as as high as that for heterosexuals.
Next, we must remember that evolutionary explanations are only needed to explain variation that has some basis in heredity. Homosexuality might have causes which are entirely non-genetic. That is, homosexuality might emerge in response to some environmental cause (or by free-will choice). Any fitness disadvantage (or advantage) would then be uneffected by selection and so be evolutionarily irrelevant. There are, of course, some problems with this explanation. No behavior is "purely" environmental in cause, so this "explanation" should raise the question why any creature would have a genetic system which permits such a seemingly-disadvantageous behavior to emerge at such a high frequency in any environment. (And for many biologists, there can be no such thing as "free will"; all behaviors must necessarily be caused by genes and environments in interaction.) But that difficulty is not insurmountable. We could, for example, resort to one of the standard limitations of evolutionary process -- maybe appropriate homosexuality-suppressing variants simply have not arisen by mutation, perhaps because the environmental stimuli that trigger homosexuality have themselves arisen too recently for evolution to have shaped appropriate counter-adaptations.
Finally, there is the adaptationist perspective. Adaptationism generally presumes that any given trait which is widespread in a population is indeed an adaptation, shaped by selection acting on hereditary variation. An adaptationist explanation for homosexuality would presume that homosexuality does have some genetic basis and that this trait is indeed associated with some fitness advantage. The question then becomes, what could that advantage be? Here the possibilities are limited only by imagination. One plausible possibility is simply that homosexuality is associated (either by direct pleiotropy or by genetic linkage) with some other trait that carries a substantial advantage. More interesting (or at least more entertaining) is the hypothesis that homosexuality itself is actually advantageous. What follows is the best suggestion I have encountered for how such an advantage might arise (at least for humans). Humans live in social groups, often including extended families. An allele for homosexuality occurred at fairly high frequency would consistently produce individuals who would not be burdened with reproductive activity of their own and who could therefore function as helpers in the community (and who might adopt specialist roles, such as teacher, artist, shaman, etc.). Groups which included such individuals might be more fit than other groups without such members. The population-genetic basis for such an explanation can be based on either kin selection or group selection. (The theoretical basis for selection favoring non-reproductive helper individuals is well established, most notably for social insects.) (Incidently, a similar style of explanation has been invoked to explain the high prevalence of mental illness in most human populations, i.e., that certain genetically-determined "special" individuals increase the fitness of their kin without reproducing themselves.)
Substantial evidence, of course, is not yet available (and is unlikely to materialize) to support or conclusively reject any of these hypotheses.
What's the point of separate sexes? If you're a hermaphrodite aren't your chances of reproducing greatly improved? ... and shouldn't that make it selected for and thus more common than it is?
This is really two questions.
- What is the advantage of outcrossing (i.e., fertilizing individuals other than yourself)?
- And what is the advantage of having male and female functions in separate individuals?
The first question, "Why outcrossing?", ought to have essentially the same answer as the question, "Why reproduce sexually?" The fact that most hermaphrodites have mechanisms to prevent (or at least discourage) self-fertilization supports the conclusion that outcrossing, like sex, must have some significant advantage(s). But (as for sex) measuring that advantage is difficult, and demonstrating that the advantage outweighs the disadvantage is harder still.
The second question, "Why have separate sexes rather than hermaphrodites?", may have a relatively straightforward answer, at least in broadest terms. The answer has two parts.
First, it is important to appreciate why there are two types of gametes -- small, active gametes (male) and large gametes well-supplied with resources for growth (female). Basically, this is attributed to disruptive, frequency-dependent selection. Assume isogamy as a starting point. Then production of either more-active gametes or better-supplied gametes should be advantageous, and the more prevalent one became the greater advantage there would be to producing more of the other. Thus we get almost universal anisogamy (sperm and eggs). (more)
Second, we ask why both types of gametes are not produced by every individual such that every conspecific would be a potential mate. Here the explanation is suggested by the fact that most flowering plants are hermaphroditic (male and female flowers on the same individual, often pistil and stamen in the same flower) while most animals are unisexual. As long as sexual reproduction can be conducted efficiently while producing both types of gametes, that condition will presumably persist. But for mobile organisms which are capable of seeking out mates (i.e., most animals), mating behavior is necessarily specialized. Now disruptive frequency-dependent selection could plausibly operate just as it does for gametes, favoring some individuals (males) who preferentially seek out multiple mates and supply minimal genetic material (sperm) and other individuals (females) who provide more abundant resources (eggs).
The text says (p. 95) that the point-mutation rate in males is higher than in females, because male cells divide more than female cells do. Is this referring only to germ line cells, or are these point-mutations occurring in DNA replication throughout the body?
This is referring to germ-line cells, since most sexually-reproducing organisms produce many more sperm than eggs (and hence need more cell-divisions to do so).
There are other sex-associated differences in mutation rate which are not well understood.
More:
304 index page
304 index pageIf the selection coefficient is as small as .001, and the initial allele frequency is low, would an observer be able to tell the difference between selection's effect and genetic drift?
With a small selection coefficient, telling the difference between selection's effect and drift is indeed difficult. If s is low enough, measuring or even detecting selection remains difficult even when allele frequency is closer to 50%, when the effect of selection on frequency is highest. Under such circumstances, measuring selection requires very large sample sizes and/or a very long periods of observation, practically unobtainable except for laboratory populations of rapidly-producing organisms (such as bacteria).
304 index pageWhat should I expect on the exam?
The exam will consist of a selection of questions posted on the study questions page. This includes not only chapter-by-chapter multiple choice questions but also the short-answer questions at the bottom of the page.
304 index pageHow does "Hardy-Weinberg" relate to drift?
The "Hardy-Weinberg" theorem describe the mathematical expectation of a stable equilibrium of allele frequences if certain ideal conditions. Those conditions include indefinitely large (effectively infinite) population size, random fertilization, absence of mutation, absence of migration, absence of mutation.
This idealized model provides a basis for considering what would happen if the idealized conditions are NOT met. Simply, genetic drift occurs in any population that is NOT indefinitely large, so that stochastic (chance) fluctuations actually matter.
(More)
304 index pageHow does genetic drift relate to variation, change, and mutation?
Drift refers to change resulting from stochastic (chance) processes, such as the "Mendelian lottery". Drift can happen only where variation exists. Drift reduces (eventually eliminates) neutral variation within a population, such that some neutral alleles are eliminated and others are fixed. New neutral alleles (new variation) are introduced by mutation. Of those new variants, some (most) are eliminated by drift, some (few) are fixed.
Over time, neutral change accumulates by drift. (Both allele-elimination and allele-fixation are examples of change.) When change by drift occurs in two or more related populations, those populations will diverge. This divergence can occur simply because of by variation among the populations in which alleles are fixed/eliminated. But this divergence is increased by the introduction of new mutant alleles into each population, some of which will drift to fixation.
(more)
In genetic drift, how are probabilities and allele frequencies related?
At an given point in time, the probability that a particular neutral allele will eventually be fixed by drift is equal to its frequency in the population. (more)
304 index pageAre there set values assigned to the strength of selection, genetic drift, and the other evolutionary mechanisms, so the net effect of all can be measured in a population?
In principle, the strength of selection and the (statistical) effect of drift can be measured or calculated for any population. However, this would require complete data on the genotype and the reproductive success of each individual in the population, so that the correlation between these could be accurately assessed. Unfortunately, such data is usually impractical to obtain. So our understanding comes from those relatively few cases where appropriate measurements can be made. In such examples, experiment and empirical measurement do support theoretical expectations.
304 index pageWhat is the difference between "density independent / density dependent" selection and "frequency independent / frequency dependent" selection?
These terms sound more complex than they really are. They are just labels to help us notice how adaptation works in particular circumstances. The circumstances are described by the words which comprise these labels.
"Density" here refers to the ecological parameter of population density. Is selection for a particular trait affected by how many individuals of the species occupy the habitat? If so, then that selection is "density-dependent".
Density-dependent selection is expected to influence certain life-style traits, in particular fecundity (maximum number of offspring produced) and parental investment per offspring. When the population density is far from carrying capacity, selection is expected favor traits which promote production of more numerous offspring. (This is sometimes called r-selection, where r refers to the rate of population increase.) When population density is near carrying capacity, selection is expected to favor traits which promote the successful survival of offspring in competition with conspecifics. (This is sometimes called K-selection (where K refers to carrying capacity.)
Both r-selection and K-selection are examples of density-dependent selection, because these modes of selection depend on population density. Any selection which is unaffected by population density is, by definition, density-independent selection.
"Frequency" here refers to the prevalence of a trait within a population. Does the contribution to fitness for a particular trait (or allele) depend on whether that trait (allele) is frequent (common) or infrequent (rare) within the population? If so, then that trait (allele) experiences frequency-dependent selection.
Perhaps the most intuitively-obvious example of frequency-dependence involves alleles which confer disease resistance. Many microbes (bacteria, viruses, other pathogens) can evolve very quickly. Once a microbe has acquired the ability to overcome the defense offered by a particular host-allele, the bearers of that allele will be vulnerable. That vulnerability is greater for alleles at high frequency (the microbe can spread more readily) than for alleles at low frequency (the microbe will have a hard time finding another suitable host).
Any selection for which the frequency of an allele does not affect the fitness of the allele is, by definition, frequency-independent selection.
(more)
304 index pageWhat do introns, exons, pseudogenes, etc. have to do with evolution?
Understanding the evolutionary role of mutation requires some appreciation for the various types of DNA changes which can occur, the various chromosomal sites for these changes, and the functional impact for each. Basically, the less critical a site is for normal function, the more likely it is that a mutation at the site will be neutral (and, hence, that neutral changes will accumulate by a combination of mutuation and drift).
- Pseudogenes basically do nothing, so almost any mutation in a pseudogene will be neutral.
Pseudogenes originate by gene-duplication mutation. At the moment of its first appearance, the "extra" copy would have no necessary function (and it might be deleterious). Unless additional mutations rather quickly established some useful role for this new gene, it could be deactivated by random mutation and would henceforward be free to vary without any impact on fitness.
- Introns are DNA segments within a gene which do not correspond to any part of the protein for which the gene codes. Although certain aspects of intron sequence do matter for normal function, details are typically less critical than for exons.
By one simplified definition, a gene is a DNA sequence which encodes a protein. However, typical eucaryotic genes have much more complex structure than this definition implies. A gene is not a simple strings of codons that specifies an amino acid sequence. Typical eucaryotic genes have complex, non-coding regulatory regions, where various regulatory factors bind to the DNA to turn the gene on and off and control its rate of expression. Furthermore, a protein is encoded by several DNA segments, called exons, which are separated by non-coding segments called introns. The entire stretch, exons plus introns, may be transcribed into messenger RNA, but the introns are then excised and the remaining spliced together before the messenger is translated into protein.
- Exons are gene segments which do code for protein, so only a very limited set of mutations (such as some third-codon-position substitutions) are predictably neutral (r nearly neutral).
The impact of base-pair substitutions which do alter a protein's amino acid sequence depends on specific features of the protein's structure and function.
- This is NOT an exclusive list. We might also distinguish many other sequence categories (e.g., telomere DNA, centromere DNA, satellite DNA, microsatellite DNA, regulatory-binding-site DNA, etc., etc...).
304 index pageAre all genetic traits correlated with reproductive success?
Simple answer: No. Although the term "genetic trait" is somewhat ambiguous, nevertheless an affirmative answer to this question would imply that there is no such thing as "neutral variation", variation which does not matter for reproductive success. In general, although it is unwise to presume that variation in any particular trait has no significance for fitness, it is also unwise to presume the opposite.
304 index pageHow do evolutionary theory and natural selection relate to human populations and human mating schemes?
Short, cheap answer: Evolutionary theory applies to human populations just the same as it does to any other species. Only the details differ (as they do for all species).
Drift happens, to the extent that neutral variation exists. Selection happens, to the extent that genetic variation correlates with reproductive success. Being human does not alter these fundamental rules. Remember, drift is just a name for change that happens according to stochastic, statistical rules. Selection is just a name for change that happens when the conditions of existence influence the passage of genes from one generation to the next. We may alter the conditions of existence, but we cannot alter the basic logic of selection. There are well-established human examples for most evolutionary phenomena.
With that said, one may note certain special features of the human species. Because much human adaptation is based on culture, stabilizing selection has probably been relaxed for many traits compared to conditions a few thousand years ago. Human "mating schemes" set the stage for sexual selection, but only in relation to human reproductive schemes. And the diversity of human behaviors sets up variously competing selection pressures, vastly complicating any effort to understand past impact or predict the future. The impact of selection mediated by infectious disease (a variety of coevolution) is probably just about as great in most current human populations as it was a few (or few thousand) generations ago.
304 index pageWhich type of selection is responsible for most adaptive features?
This question is interesting but probably unanswerable. The different "types" of selection are labels given to help us understand the various ways in which selection can operate. The textbook listing (see notes) is not exclusive. One might label many other "types" of selection, as the occasion arises (e.g., cold-climate vs. warm climate selection, or predator vs. predator-free selection, or low-salt vs. high-salt selection, etc., etc...).
There is no objectively meaningful way to weigh such categories in the balance for comparison, just as there is no good way to count "adaptive features". Furthermore, any given trait may respond to different types of selection at different times. For example, a trait initially shaped by directional selection may be maintained by stabilizing selection.
Nevertheless, one may note that each "type" of selection is as prevalent as the circumstances through which it operates.
- For established populations experiencing reasonably stable environments, current adaptation is probably sustained largely by stabilizing selection (i.e. departures from the prevailing optimum will be selected against). In populations that have recently experienced environmental change (including various modes of coevolution), directional selection may be important (i.e., prevailing trait values may not be optimal in the changed circumstances). Disruptive selection depends on rather special conditions and may therefore be less prevalent (although disruptive selection may be especially important during sympatric speciation).
- Sexual selection is a special case of selection-in-general (such that "natural selection" refers to all selection except for "sexual selection"). Thus sexual selection affects only those traits which influence fitness through mating success. All other traits would (by definition) be subject to "ordinary" natural selection.
- Density-dependent selection affects only those traits for which influence fitness in relation to population density. All other traits would (by definition) be subject to density-independent selection.
- Similarly, frequency-dependent selection affects only those traits whose impact on fitness depends on the frequency of the trait within the population. All other traits would (by definition) be subject to density-independent selection..
304 index pageWhat is coevolution?
"Coevolution" is the name given to the process of evolution when two (or more) species evolve in relation to one another, such that evolutionary change in each species mutually affects the evolution of the other.
Although coevolution receives little attention in our text, it is a very interesting and important topic in evolutionary biology. Because other species are such significant elements in any species' environment, much (most?) adaptation probably involves some degree of co-evolution. However, special attention (and application of the label) is usually reserved for the more extreme cases of closely-associated pairs of species, including various modes of symbiosis (mutualism, predator-prey and parasite-host arms races, etc.).
304 index pageWhat causes genetic change?
The causes of genetic change are manifold. We must pin this question down a bit.
By one definition, evolution is changes in the genetic composition of populations. Thus this question asks for nothing less than an explanation of the mechanisms of evolution. The short answer is simply a list of all those processes which can result in changes in the genetic composition of a population.
- drift
- bottlenecks
- founder effects
- selection
- migration
- mutation
For all of these except mutation, change can result only when there is pre-existing variation. Mutation is the process which introduces variation in the first place. The process of mutation has many causes, involving chemical details of DNA replication, proofreading, and repair. The topic of mutation lies beyond the scope of this course.
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