BIOL 606 Home

Measuring Fitness in Natural Populations with Particular Reference to Natural Selection

Lecture by Jan Jekielek

Rappateur: Renee Polziehn

In Origin of the Species, Darwin recorded observations that any variation in a trait which provides an advantage to an organism, no matter how slight, will prolong its life. If the character is heritable, the offspring will also have a greater likelihood of surviving. Over time, the proportion of individuals with the selective advantage will increase. Therefore, fitness is a measure of the process of natural selection and the potential for evolutionary change is found in the variation of survival and reproduction.
Futuyma (1998) described fitness as the average lifetime contribution of an individual of a particular genotype to their population after one or more generations. Absolute fitness is a measure of the total number of offspring produced by an organism in its lifetime and is related to the adaptedness of the organism to the environment. The relative fitness or Darwinian fitness is the average reproductive success of one genotype relative to the other genotypes in the population, where the group of interest is given a fitness of 1.0 and other groups are graded against it. Indirect fitness is a measure of the benefit an individual gains when its related kin/offspring are successful. Inclusive fitness is extensively used to study the benefits of social behavior, and is described as Darwinian fitness plus the advantages gained from interactions of an individual and its neighbors. In social insects the fitness of each individual depends on the fitness of the entire population.

Selective forces act on all stages of an organisms' life, and fitness in the prezygotic stage depends on gametic viability and fertilization success, while fitness in the post-zygotic stage is determined by mating success, fecundity, offspring survivorship, and reproductive lifespan. Variation in behavioral, morphological, and physiological traits that are thought to affect the fitness of an individual are measured and correlated to fitness components. The traits can be polymorphic or continuous, and must be heritable. Cross-fostering experiments (as employed by Schluter and Smith 1986) which find a positive correlation between traits shared by an offspring and its true parents are heritable, while those between an offspring and its foster parents are due to environmental influences. Ritland (1996) found testing heritability in a lab was less likely to find a positive correlation than testing in the field. Perhaps the range of environmental variables produced in the natural setting tested genotypes more thoroughly than in the laboratory.
Genetic markers were used to determine which generations are under selection in a population of eelpout (Zoarces viviparus ). Homogeneity of both sexes over all age classes was observed, and the allelic frequencies for the EST III protein supported random mating and an equal production of all gametes. However, sampling of the offspring over four years detected a deficit of heterozygotes which was caused by selection against juvenile heterozygotes.

Many experiments try to determine which traits are being selected upon and the biological significance of the selection. Schluter and Smith (1986) looked at six heritable traits over four seasons to determine if fitness differences were associated with morphological traits in song sparrows (Melospiza melodia ). Selection favored juvenile females with longer beaks and shorter tarsi, and the opposite in adult females. A path analysis identified the traits which contributed to fitness and the intensity of selection for each trait at different life stages. Different body types were also found to excel in different seasons for the medium ground finch which led Gibbs and Grant (1987) to estimate lifetime fitness as a net effect of the different selections of a phenotype over a lifetime. Studies using fitness based on only one portion of the lifecycle or a few traits may not produce a reliable estimate of the total selection acting on the fitness components. Short term evolution, therefore, is more accurately predicted when selection gradients are calculated by regressing total lifetime fitness over a number of phenotypic traits.

Selection differentials relate change in a trait and a fitness component, and may be either stabilizing or directional. Selection gradients for a given trait represent the slope of the relationship between fitness of that trait after removing correlations with other measured traits, therefore, indicating the intensity of selection for a trait. Multiplicative fitness components (MFC) are fitness components broken down for each step of the lifespan of an organism (e.g. number of copulations/year/lifespan). Lifespan is broken down into biologically meaningful units and selection is measured for each one. However, MFC will not indicate which component is most important for total lifetime fitness. Conner (1996) combined path analysis which provides correlations similar to the selection gradients for individual fitness components and MFC, where regressions of each fitness component over total lifespan indicate the component most important in total fitness. This method allows the complex interactions of an organism, the environment, a number of phenotypic traits, and multiple fitness components to be portrayed in a clear and concise manner, and can highlight relationships that may otherwise be overlooked.

Conner, J.K. 1996. Ethology Ecology and Evolution : 387-397 Ritland, C.K. 1996. Evolution 3: 1074-1082
Futuyma, D. 1996. Evolutionary Biology. S innaur Associate Schluter, D. and Smith, J.N.M. 1986. Evolution 40: 221-231
Gibbs, H.R., and Grant, P.R. 1987. Nature 327: 511-512


Discussants: Ranessa Cooper and Chris Kyle
Rappateur: Renee Polziehn

The following points were made in discussion of the concepts and methods used by Conner (1996) and Schluter and Smith (1986):

1) Song sparrows were studied over the period of 1975-1979. An intensive amount of sampling was performed during this period. Do studies analyzing fitness and natural selection require this amount of work? Most studies are performed over two years and conclusions can be drawn for only a short time period. Long term studies will show different effects and become both more useful and difficult to interpret. Smith is able to use the song sparrow as a long term model for selection and fitness experiments because students and colleagues continue to collect data from the same population.

2) Selection gradients, multiplicative fitness component analysis, and path analysis have been used for many years. Has Conner (1996) added anything new and are there advantages of his approach? Multiplicative fitness component analysis has been used since 1984, but Conner has provided new ways to visualize all the interacting components. Perhaps the inability of MFC to identify which fitness components are most important to the overall fitness of an organism should have been stressed. Conner's combination of path analysis and MFC leads to a greater understanding of which traits are under the most selection and have the most impact on total fitness. Path analysis was also shown to be a more concise way to get across a lot of information and relationships were more obvious than data listed in selection gradient tables.

3) Could Schluter and Smith (1986) have added anything to improve their paper? There were no quantitative measurements of the phenotypes or information on the fecundity and juvenile survival for different selection at different lifetimes. The effect of the two fitness components on total lifetime fitness would have been interesting, because it is possible they would have canceled each other.

4) Would this method help to make comparisons amongst taxa? Comparisons can only be drawn if both studies compare the same traits. If the method is a good method, then there should be an increase in the use of the technique. The technique has the ability to highlight the unexplained, and identifies the real character that needs to be measured.

5) How do you choose the particular traits to measure fitness? Characters are usually chosen which have an impact on fitness components, such as lifespan and mating success. Each component is examined for traits which can be measured, such as the number of inseminations per female. Length of horn and inseminations were easy to measure but are not extensive. The number of inseminations for male insects is one of the better estimates of male fitness because females can store sperm from many males over a long time period. Horn length, size, and weight also show a positive correlation with fitness which make it a good character. By looking at the 'p' values, one can determine whether a character has a significant impact on fitness components and the overall fitness.

6) What is the difference between direct and indirect fitness? Fitness defined by Schluter and Smith (1986) was the number of offspring surviving to independance. More commonly, fitness is defined as the ability of an individual to produce offspring which themselves produce offspring. In the former situation, selective advantages such as the large breast stripes in the male offspring of Great tits would be an indirect benefit to the female's fitness. In the latter definition, the same event would be direct fitness. If fitness was defined by allelic extinction, which is the time required for an allele to disappear from the population, selective advantages to fitness could be derived for many generations.

8) What do you do with a value after you calculate fitness? The fitness values may provide an indication if selection is directional or stabilizing in the short term. Because most fitness values are based on only a few years of selection, the direction of selection over the long term cannot be accurately predicted.

9) Is anyone measuring fitness? Rich Palmer suggested some aspects of the Conner method may be applied to a study of smooth versus sculptured shells of mullusks. These organisms show a pronounced cline over 5-6 km, where predominantly smooth forms are found in rough wave activity and sculptured forms are found in gentle wave activity. Lab experiments produce two distinct phenotypes, while a range of phenotypes can be found in natural environment. Direct fitness cannot be measured, but relationships such as the abscence of sculptured forms and survival or predation will be estimated. A greater plasticity in heterozygotes is thought to make these forms more adaptable to change and may indicate why both types exist.

BIOL 606 Home