For her dissertation research, Kerry Bright (Duke University) has been examining the type of natural selection, if any, that acts on leaf-shape variation in the annual morning glory, Ipomoea hederacea.  This species grows throughout much of eastern North America.  Throughout the southern portion of its range, populations often exhibit discrete variation in leaf shape.  This variation is controlled by underlying variation at a single locus.  Individuals that are homozygous for one allele, which we will designate here as A, have leaves that are strongly lobed.  Heterozygotes are slightly less strongly lobed, while heterozygotes for the other allele, which we designate here by a, are unlobed, or entire (see Elmore, C. S. 1986.  Mode of reproduction and inheritance of leaf shape in Ipomoea hederacea.  Weed Science 34: 391-395).

     Kerry estimated the fitness of these three genotypes in two different experiments.  One experiment was conducted at a site near Durham, NC (Bahama site), at which the native population of morning glories is nearly fixed for the A allele.  Most of the plants in this population are thus the strongly lobed AA genotype.  The second experiment was conducted at two sites in southeastern NC (the Whiteville and Clinton sites).  The populations at these sites are polymorphic for leaf shape.

     To measure fitnesses at each site, Kerry generated experimental seeds of known genotype at the leaf-shape locus.  For example, to generate AA seeds, she crossed a known AA seed parent with a known AA pollen parent in the greenhouse.  Similarly, to produce Aa seeds, she crossed AA with aa, and to produce aa seeds she crossed two aa individuals.

     The experimental seeds were then planted out at the study sites in June.  Emergence of seedlings and their survival was monitored weekly, and when plants started producing seeds, all seeds were collected and counted for each plant.  These two experiments yielded the following data:

 Bahama Site                             AA               Aa                aa

 No. seeds planted                      320              320              320
 No. seeds surviving to                290              287              294
  produce seeds
 No. seeds/surviving plant           11.6               9.6               7.6    *
 Survival Probability                  0.906            0.897           0.919
 W                                             10.5               8.6               7.0    *
 
 

 Whiteville Site                          AA               Aa                aa

 No. seeds planted                       186              186              186
 No. seeds surviving to                 172              170              176
  produce seeds
 No. seeds/surviving plant            98.2            115.6            98.6   *
 Survival Probability                   0.925            0.914          0.946
 W                                              93.3            108.7            93.3   *
 

 Clinton Site                               AA               Aa                aa

 No. seeds planted                       155              155              155
 No. seeds surviving to                 131              134              132
  produce seeds
 No. seeds/surviving plant             94.5           113.4             93.9  *
 Survival Probability                    0.845           0.865           0.852
 W                                               79.9              98.1            80.0  *
 
         * indicates statistically significant difference among genotypes

    Two components of fitness were estimated in these experiments: (1) survival probability and (2) Number of seeds produced.  Because this species is highly selfing (less than 5% of seeds are pollenated by pollen from another plant), the male and female components of fitness are both reflected primarily by the number of seeds produced.  In an annual like Ipomoea hederacea, overall fitness is just the product of these two fitness components, i.e.

             W = (survival probability) x (No. seeds produced by a surviving plant)

    At the Bahama site, while the estimated survival probabilities for the three genotypes were not exactly the same, the small differences among the estimates can be explained by chance effects.  A proper statistical analysis revealed that there was no detectable real difference in survival probability among the genotypes.  By contrast, the differences in number of seeds produced by the three genotypes at this site are too large to be accounted for by chance effects.  A proper statistical analysis of the data indicates that the actual fitness rankings in this experiment are W_AA > W_Aa > W_aa .  Based on these rankings, we would expect the allele to be fixed at this site, which is what is seen.

    At the Whiteville and Clinton sites, the small differences in survival probability among genotypes are again attributable to chance effects.  There are no statistically significant differences in survival among the genotypes.  At both sites, however, number of seeds produced per plant differs significantly among the genotypes.  In particular, at both sites, the heterozygote has the highest fitness, i.e.  W_AA < W_Aa > W_aa .  Based on these rankings, we would expect a polymorphism to be maintained at these sites, which is what is seen.

    It is not yet clear what environmental factors cause the fitness differences among genotypes, or why the fitness rankings differ between the Bahama site, on the one hand, and the Whiteville and Clinton sites, on the other.  Kerry is currently examining the effects of herbivores, pathogens, and heat and water stress in generating fitness differences among genotypes.Genetic makeup

            [ Back to Lecture 4 | Lecture Outlines | Bio 120 Home Page | Department of Biology | Duke University ]