Principles of Evolution Web Page
                        Lectures
 



LECTURES 11 and 12:    ALTRUISM AND KIN SELECTION 

I.  Introduction

    A.  Natural selection and evolutionary change.

         1.  We have seen that natural selection is the single most important process responsible for evolutionary change, and
               we have developed a conceptual famework for understanding evolution by natural selection in natural populations

         2.  We have seen that natural selection produces adaptation by eliminating less fit genotypes and promoting the
              persistance of alleles conferring high fitness.

         3.  The usefulness of the conceptual framework we have developed lies in its general applicability.  Like any theory,
              a prediction of this framework that runs counter to known facts is likely to make us question the generality of
              that framework.

         4.  In general, the theory of evolution by natural selection is compatible with most of the patterns and phenomena we
              see in nature.   Even Darwin realized, however, that there are certain cases of adaptation in nature that appear to
              run counter to the theory of evolution by natural selection.

         5.  In this lecture we examine these cases and expand our understanding of natural selection to accomodate these cases.

II.  A specific example: Sterile castes in insects

    A.  For Darwin, one of the major challenges to his theory of natural selection lay in explaining the evolution of sterile castes
          in social insects.

        1.  It was known long before Darwin that many insects exhibit a high degree of social organization.

        2.  For example, many species of termites, ants, bees and wasps form colonies composed of several hundreds
             of thousands if not millions of individuals.

        3.  Of these individuals, only one or occasionally a few are reproductive.

        4.  For example, consider the army ant Eciton burchelii.

            a.  Queens--lay all eggs of colony; have greatly modified morphology: they are essentially egg machines, capable
                 of producing 100,000 to 300,000 eggs in 2-week period, millions over a lifetime.
            b.  workers--sterile; all female; forage for food for both queen and immature ants; take care of immature ants; etc.
            c.  soldiers--sterile; all female; protect nest
            d.  there is thus reproductive division of labor within the colony.

    B.  The evolutionary problem

         1.  All ants live in societies and exhibit reproductive division of labor.

         2.  Ants evolved from solitary species that did not exhibit reproductive division of labor.

         3.  At some point in the ancestral species that gave rise to ants, then, alleles that caused individuals to form societies
              and that caused workers to be sterile arose and were fixed by some process.

         4.  For Darwin, and many successors, it was hard to understand how that process could be natural selection.

         5.  In particular, it was difficult to understand how sterile individuals could pass on alleles that rendered them sterile to
              the next generation, and this how the frequencies of those alleles could increase in a population.

         6.  Darwin never came up with an answer to this dilemma; in fact, nobody did until the mid 1960's.

         7.  We will see shortly what the intellectual breakthrough was at this time; first, however, let us consider some
              other examples of this phenomenon
 

III.  Other examples of Altruism

    A.  The phenomenon of Altruism

         1.  Sterile castes of insects are just one example of a whole class of traits, collectively known as  altruism.

         2.  All altruistic traits pose the same difficulty for the theory of natural selection, and thus we must have some means
              of reconciling the general phenomenon of altruism with our conceptual framework if we wish to preserve that
              framework.  It is therefore crucial that we know exactly what we mean by altruism.

         3.  An altruistic trait is one that enhances the ability of another individual to survive and/or reproduce while at the
              same time it decreases the ability of the individual possessing the trait to survive and/or reproduce.

         4.  In other words, an altruistic trait decreasese the fitness of the individual posessing the trait (we thus say there is a
              cost to altruism) and increases the fitness of one or more other individuals (we call this the benefit of altruism).

         5.  The individual possessing the altruistic trait is called the donor, or, simply, the altruist.  The donor's fitness is
              lowered by the altruistic trait.

         6.  The beneficiary of the altruism is called the recipient.  The recipient's fitness is enhanced by the altruistic trait.

         7.  In the case of Eciton,

              a.  sterile workers and soldiers are donors
              b. queens are recipients
              c.  the altruistic trait is a complex series of behaviors that include forgoing the ability to reproduce, gathering food
                   for the queen and her offspring, defending those offspring and the queen from predators, and so on.
              d.  Clearly, by being sterile, workers have reduced fitness--in this case their fitness is 0.
              e.  At the same time, the queen, as recipient, has a greatly enhanced ability to reproduce--she could not possibly
                   lay as many eggs as she does if she did not have all her needs tended to by a sterile caste of workers.

    B.  Other examples of altruism.

         1.  Other social insects

              a. Sociality, with sterile castes, has evolved independently on the order of 15 different times in the Hymenoptera
                  (ants, bees, and wasps).
              b. Sociality and reproductive division of labor has also evolved in termites

         2.  Naked mole rats have evolved a social system similar to that of social insects; within a colony, there is single
              reproducing female (queen) and a number of non-reproducing workers.
 

         3.  Portugese man of war and related coelentrates

              a.  What looks like a single individual is actually a colony of individual animals
              b.  Several types of individuals
                  i.  float
                  ii. gastrozooid (digestion)
                  iii. dactylozooid (food capture and defense)
                  iv. gonophore (reproduction)
              c.  In this colony, individuals that specialize as float, gastrozooids, and dactylozooids have completely given up
                  ability to reproduce.
              d.  At the same time, because all these individuals are organized into a highly integrated colony, the reproductive
                   ability of the gonophores is greatly enhanced over what it would be for an individual if it did not live as part of
                   a colony.

         4.  Florida Scrub Jays (Aphelocoma)

              a.  Males and females form pair bonds and set up a breeding territory in spring.
              b.  Often, there is an additional individual that joins the pair--a helper.
              c.  Helper does not perticipate in nest construction or incubation of eggs, but is active in just about every other
                   activity including food gathering, eeding of the young, defense of the territory against intrusion by other jays,
                   and protecting the young from attacks by predators.
              d.  Helper stays with the pair through the entire breeding season and thus forgoes the opportunity to reproduce
                   that year, i.e. it reproduces its own reproductive capacity by helping.
              e.  A field study by Woolfenden demonstrated that--assistance provided by helpers increase reproductive success
                  of the helped pair:
                    i.  The study banded newborn birds from all nests in a population for five years and determined the number
                        of offspring surviving to leave nest for 121 different nests.
                    ii.  Of these, 74 nests had helpers.
                    iii.  Average number of  young birds surviving to leave nest was 2.1 in nests with helpers, 1.1 in nests
                         without helpers.
                    iv.  Additionally, only an average of 0.5 offspring were alive three months later for nests without helpers,
                         while 1.3 were alive from nests with helpers.
                    v.  Helpers thus clearly are altruists--they reduce their own fitness while increasing that of the pair they assist.

         5.  Alarm calls

              a.  Many animals, such as squirrels, give alarm calls when predators approach.
              b.  In response to alarm calls, other individuals take appropriate action to avoid predator.
              c.  Alarm calls clearly benefit the animals that hear themn by decreasing predation and increasing survival.
              d.  There is also some evidence that the animal giving the alarm call may suffer to some extent by doing so.  In
                   giving the call, the animal may attract the predator's attention and make itself more vulnerable to predation.
              e.  Alarm calls are thus thought to be altruistic.

IV.  The General Problem with Altruism

    A.  The problem with an evolutionary interpretation of altruism, based on all that we have seen so far, is that
          altruism apparently should not evolve, at least according to our conceptual framework.

         1.  To see this, let us assume that altruism is controlled by a single locus with two alleles.

         2.  This assumption is admittedly probably overly simplistic, but one can make an argument similar to the following
              even for altruistic traits that exhibit more complex genetic control.

         3.  Consequently, to provide a basic understanding of the problem altruism poses to evolutionary theory, we
              will concentrate on the one-locus case.

    B. Genetic model of the evolution of altruism

         1.  When directed randomly to all potential recipients, altruism can not evolve.

         2.  However, when directed only to other potentially altruistic individuals, altruism can evolve if the benefit of
              altruism exceeds the cost of altruism.

         3.  When this criterion is satisfied,  the increase in the number of copies of the altruistic allele produced by  recipients
              is greater than the number of copies that are lost due to reduced  reproducion by the donor.
 
         4.  These considerations thus demonstrate that the evolution of altruism is not theoretically prohibited.
 

V.  Hamilton's rule

    A.  It is probably unlikely that an altruist would be able to detect another individual's genotype at loci influencing
          tendency toward altruism and thus be able to preferentially direct altruism toward specific genotypes, as is required
          by the previous argument.

    B.  However, such discrimination is not necessary for altruism to evolve.

         1.  Instead, if altruism is directed preferentially toward close relatives, it can still evolve.
 
         2.  Since it is known that many organism can distinguish close relatives from other individuals, this provides a
              possible avenue for the evolution of altruism.

         3.  This mechanism--the generation of a type of natural selection by preferentially directing altruism toward close
              relatives (kin)--is called kin selection.

    C.  How kin selection works.

         1.  Formally,  we are trying to account for how a mutant allele that causes altruism, initially at low frequency, can
              increase in frequency in a population.

         2.  To understand this, we can conceptually divide any change in the frequency of allele a for altruism into
              three components:

                Dpa = (1)  increase in p  due to altruism directed toward aa and Aa individuals    +

                           (2) decrease in due to altruism directed toward AA individuals              +

                           (3)  decrease in p  due to detrimental affects of altruism on donors (cost of altruism)

              a.   Case 1: altruism is directed randomly
                  i.   Then effect (1)  effect (2) are equal and cancel each other.
                  ii.  The net effect is then due entirely to effect (3), and the frequency of altruism will decrease, as our original
                       model predicted.

              b.  Case 2: altruism directed only toward other aa individuals

                  i.  Then effect (2) is zero.
                  ii.  Consequently, Dpa > 0, and the frequency of altruism increases, as long as effect (1) is greater than effect  (3).
                  iii.  The handout shows this will occur when the cost of altruism is less than the benefit.

               c.  Case 3: altruism directed toward close relatives

                  i.  Close relatives, because they tend to share the same genes, will also tend to be aa and Aa individuals (aa
                      are potential altruists).
                  ii.  Consequently. altruism will be directed preferentially toward aa and Aa individuals, though not exclusively.
                  iii.  This means that effect (1) > effect (2),
                  iv.  which means that the joint effect of (1) and (2) is an incremental increase in pa
                  v.  If this incremental increase is large enough, it will be greater than effect (3), and the net effect will be an
                       increase in the frequency of a.
                  vi.  W. D. Hamilton determined the conditions under which this will be true:
                  vii.  Hamilton's rule    b/c  > 1/ r  ,
                  viii.  where r is the average degree of relatedness of all recipients to donor (see table on handout).
                  ix.  b is the benefit of altruism, given by

                            (fitness of helped individual) - (fitness of unhelped individual)

                  x.  and  c is the cost of altruism to the donor, and is

                            (fitness of unhelped individual) - (fitness of altruist)
 

        4.  Hamilton's rule reveals three important things about the evolution of altruism:

            a.  First, it shows that there are conditions under which altruism is favored by natural selection; consequently,
                 it reconciles altruism with our conceptual framework of natural selection.
            b.  Second, it shows that it is much easier for altruism to evolve when it is directed toward closely related individuals.

                i. Closely related individuals have high r values, so that 1/r is small, and hence the benefit conferred by an
                   altruist need not be exhorbitantly great (measured in units of c).
                ii.  For distant relatives, by contrast, r is small and 1/r is therefore large.  In the extreme case of r = 0, 1/r is infinite
                    and no value of b will satisfy the criterion.  Consequently, the evolution of altruism is impossible, as we saw earlier.
 
            c.  Third, it leads to the prediction that when altruism is observed in nature, it should be directed toward close
                 relatives.  This is a prediction that can be tested empirically.

VI.  Tests of prediction: Examples

        A.  Florida Scrub jays

        1.  Remember that Woolfenden found that some mating pairs are joined by auxiliary helpers that do not reproduce,
             but that contribute to feeding the young, defending the nest, and so on.

        2.  By marking young birds in the nest and then following them over a period of several years, he was able to
            determine the relatedness of helpers to the individuals they helped.  Of the 74 cases of assistance he observed,
            he found that

                48 assisted both parents
                16 assisted a father and a stepmother
                  2 assisted a mother and a stepfather
                  7 assisted a brother and his mate
                  1 assisted a totally unrelated pair.

        3.  Thus, Florida scrub jays direct their altruism toward very close relatives, as predicted by Hamilton's rule.

    B.  Polistes wasps (M. J. West-Eberhart)

        1.  Paper wasps in the genus Polistes exhibit two types of altruism
 
            a.  In the spring, when overwintering inseminated females are starting new nests, reproductive primary foundresses
                 will be joined by secondary foundresses.  The ovaries of the secondary foundresses regress, and they become
                 non-reproducing workers assisting the reproductive primary foundress.

            b.  Offspring of the primary foundress (except for reproductives produced in the fall) are sterile workers.

        2.  In the fall, West-Eberhart marked reproductives from a large number of nests.  Individuals from a given nest
             were given marks that would allow them to be distinguished from individuals from other nests.

       3.  The following spring, West-Eberhart determined the identities of primary and secondary foundresses in the new
            nests being formed.

       4.  She found that in most cases, secondary foundresses were from the same nest as the primary foundress they helped,
            i.e. secondary foundresses were helping sisters.
 

VII.  Conclusions

    A.  Altruism is a common phenomenon in nature.

    B.  Altruism can evolve by kin selection

        1.  At first sight, altruism appears to run counter to ideas about how natural selection operates, since it involves
             the reduction of an individual's reproductive ability while enhancing that of another's.

        2.  In extreme cases, such as the sterile castes of social insects, altruistic individuals do not reproduce at all, and it
             was difficult to understand how they could pass on genes for altruism to succeeding generations.

        3.  Upon closer analysis, however, it is found that the key to understanding the evolution of altruism lies in two
             important facts:

            a.  Altruistic individuals tend to direct their altruism toward close relatives
            b.  Those close relatives will tend to carry copies of the altruistic alleles.
            c.  These facts mean that, if the benefit conferred on recipients is sufficiently large, compared to the cost of altruism,
                 the frequency of the altruistic allele will increase, i.e. altruism will evolve.

 .  


    [Lecture Outlines | Bio 120 Home Page | Department of Biology | Duke University ]