Demonstrating and measuring evolution in a laboratory environment with animal species is typically very challenging. However, insects are among the most amenable species for such studies and the bean beetle,Callosobruchus maculatus, is the easiest species in which such studies may be conducted. Bean beetles are agricultural pest insects of Africa and Asia. Females lay their eggs on the surface of beans (Family Fabaceae) and the entire pre-adult parts of the life cycle occur inside the host bean. In this study, students design and conduct experiments to evaluate whether evolution by natural selection (or alternatively, genetic drift) may be induced in laboratory populations.
Topic: Evolution and Natural Selection
Level: Introductory majors – Upper-level majors
Class time: one 2-3 hour class period, plus 1 hour class period two weeks later, and one 2-3 hour class period about 4 weeks later
Lawrence S. Blumer1 and Christopher W. Beck2
1Department of Biology, Morehouse College, Atlanta, GA 30314
2Department of Biology, Emory University, Atlanta, GA 30322
Objectives
Design and perform an experiment in which either evolution by directional selection (or genetic drift) is induced in laboratory populations of bean beetles, Callosobruchus maculatus.
Evaluate control and experimental populations to measure evolutionary change.
Introduction
Evolution is defined as genetically based phenotypic change that occurs over generational time spans. Although natural selection is typically the most potent cause for evolution and is the principal cause for evolutionary change, other processes, such as mutation, migration, and genetic drift also can cause evolution (Freeman and Herron 2007). Natural selection occurs when a trait, such as body size, varies in a population and individuals differ in their survival and reproductive success as a consequence of the particular character of a trait. For example, if adult body mass varied in a population and the risk of predation were greater among the smallest individuals in the population, then the larger individuals would have greater survival and consequently greater reproductive success than the smaller individuals. Directional selection, such as that described for body mass, may result in directional evolution if the variation in the trait (body mass variation in this example) were caused by genetic differences among individuals. In other words, if the variation in body mass were heritable, then directional selection on body mass would cause directional evolution in body mass. The other potential causes for evolution (mutation, migration, and genetic drift) are real, but all cause random phenotypic changes in a population. Mutation is the spontaneous change in the genotype of an individual that may cause a change in the phenotype of the offspring of that individual. Migration is the movement of individuals into a population (immigration) or movement out of a population (emigration). In either type of migration, the mean value for a trait and variation in that trait in a given population may change as a result of such movement. Genetic drift results from random genotypic change that can occur when a population has very few individuals among whom reproduction occurs. When a population contains few individuals, even random mating may result in the loss of alleles and an increased frequency of homozygous genotypes compared to populations with greater numbers (Futuyma 1986). Genetic drift is a form of reproductive sampling error. To the extent that random changes in genotype frequencies result in changes in phenotypes, phenotypic evolution may occur as a consequence of genetic drift.
In this study, you will design and conduct experiments to induce evolutionary change in an insect species, bean beetles (cowpea seed beetles), Callosobruchus maculatus. Bean beetles are agricultural pest insects of Africa and Asia. Females lay their eggs on the surface of beans (Family Fabaceae). Eggs are deposited (=oviposition) singly and several days after oviposition, a beetle larva (maggot) burrows into the bean. At 30°C, pupation and emergence of an adult beetle occurs 25-30 days after an egg was deposited. Adults are mature 24 - 36 hours after emergence and they do not need to feed. Adults may live for 7-10 days during which time mating and oviposition occurs (Mitchell 1975). Adult body mass, linear body dimensions, and egg-to-adult development-time are variable traits in C. maculatus. Consequently, these easily measured traits are candidates for inducing evolutionary change in laboratory populations. Previous studies have found that variation in body mass is heritable in both sexes but failed to find heritable variation in egg-to-adult development-time (Fox et al 2004).
Experimental Design
Your instructor will announce the focus of your study, either inducing evolution by natural selection or genetic drift. Address the following questions. Come to class ready to discuss your answers.
- What is the importance of trait variation if you were to induce natural selection?
- Design an experiment or set of experiments to test the hypothesis that you can induce evolution in a trait by natural selection or genetic drift.
- What assumption must you make about the cause for trait variation if selection were to result in evolution? Can natural selection result in evolution in a trait if variation were caused entirely by environmental variation?
For each of the experiments you designed above, you should:
- predict the possible outcomes for the experiment that would support your hypothesis
- identify and list the variables you would manipulate in your experiment
- identify and list the variables you would keep constant in your experiment
- list the data you would collect to determine if your predictions were true
Literature Cited
Fox, C.W., M.L. Bush, D.A. Roff, and W.G. Wallin. 2004. Evolutionary genetics of lifespan and mortality rates in two populations of seed beetles, Callosobruchus maculatus. Heredity 92:170-181.
Freeman, S. and J.C. Herron. 2007. Evolutionary Analysis. 4th edition. Benjamin Cummings. 800 pages.
Mitchell, R. 1975. The evolution of oviposition tactics in the bean weevil, Callosobruchus maculatus F. Ecology 56:696-702.
Futuyma, D.J. 1986. Evolutionary Biology. Second edition. Sinauer Associates. 600 pages.
This study was written by L. Blumer and C. Beck, 2008 (www.beanbeetles.org).
Copyright © by Lawrence S. Blumer and Christopher W. Beck, 2009. All rights reserved. The content of this site may be freely used for non-profit educational purposes, with proper acknowledgement of the source. All other uses are prohibited without prior written permission from the copyright holders.
The student handout is written as a guided inquiry that allows students to design their own experiments, rather than instructors giving students explicit directions on how to conduct their experiments. No matter the exact experiment that students design, the experiments will require having dense cultures of bean beetles from which females can be isolated. If new cultures are initiated approximately 2 months before the lab period, there will be sufficient time for two generations of beetles, which will result in dense cultures. When possible, we supply one culture to each group of students. However, cultures should have sufficient beetles for multiple groups. Newly emerged cultures work better for this experiment than older cultures.
Instructors should caution students to prevent the accidental release of bean beetles from the laboratory environment. Callosobruchus maculatus is a potential agricultural pest insect that is not distributed throughout the United States and Canada. It is essential that you keep your cultures secured in a laboratory environment to ensure that they are not released to the natural environment. Disposal of cultures (and beans (seeds) exposed to live beetles of any life cycle stage) requires freezing (0°C) for a minimum of 72 hours prior to disposal as food waste. If you have any questions about the handling or disposal of bean beetles, please contact Larry Blumer at lblumer@morehouse.edu or 404 658-1142 (voice or FAX). Information also is available at: www.beanbeetles.org in the Handbook section.
Experimental Design
This study is different from most experiments in that we are asking students to design an experiment to induce a predicted outcome and assess whether that outcome occurred. Students will test the hypothesis that they can induce evolution in bean beetles. The most obvious traits that are highly variable and readily measured are body mass and development time (time from egg-to-adult emergence). Two types of experiments are possible:
- A directional selection experiment in which one extreme phenotype is selected to start a new population or
- A genetic drift experiment in which a population bottleneck is created.
Directional Selection This experiment could be performed by students selecting one extreme phenotype to start new populations. The extreme phenotype could be at either end of the trait distribution, so either up-selection or down-selection is possible. For example, the variation in body mass of pre-experiment control populations could be evaluated so that individual males and females with body mass in the top 5% of the distribution (up-selection) [or the bottom 5% of the distribution (down-selection)] could be selected to be founders of a new population. Similarly, the distribution of development time could be evaluated in pre-experiment control populations so extreme selection criteria could be proposed. We suggest that new populations should be started with 15 to 25 males and females (using the same numbers for both selection treatments and control populations). New control populations should be run simultaneously with selection treatment populations to ensure that any changes seen in the selection treatments are a consequence of selection and not an environmental effect. Since development can be completed in as few as 3 or 4 weeks at 30°C, several generations of a directional selection experiment could be performed in a semester course. However, data collection requires that students attend to cultures every day once adult beetles begin emerging. Data on development time requires a record of the date each adult emerged, so each day all newly emerged adults would need to be collected and removed from each culture. Similarly, data on body mass requires that beetles are weighed within 24 hours of emergence because adults do not feed and body mass decreases with age and as a consequence of mating (for males) and egg laying (for females). Data collection should occur for 7-10 days once adults begin emerging. Having each pair of students run one control and one selection treatment replicate culture will minimize the amount of data collection each student must perform while ensuring adequate replication in a class of 20-30 students. Conducting a selection experiment to induce evolution requires the assumption that the observed variation in the trait being selected is caused by genetic differences between individuals (variation must be heritable). For example, we assume that parents with greater than average body mass will produce offspring with greater than average body mass. Egg-to-adult development has very small or zero heritability so even very extreme selection is unlikely to yield evolutionary change.
Genetic Drift This experiment is simpler than a selection experiment since the characteristics of the treatment populations need not be quantified at the start of the experiment. Students could start control and genetic bottleneck treatment cultures by manipulating the number of adult beetles that start a new culture. Our students have done this with control cultures of 15 or 25 randomly chosen males and females and bottleneck cultures of 3 or 5 randomly chosen males and females. As in the selection experiment, the control and bottleneck treatments must be run simultaneously to ensure that any changes observed in the bottleneck treatments are due to drift and not environmental effects. Once adult beetles begin to emerge after 3-4 weeks data must be collected each day for 7-10 days from every culture. Data on development time requires a record of the date each adult emerged, so each day all newly emerged adults would need to be collected and removed from each culture. Similarly, data on body mass requires that beetles are weighed within 24 hours of emergence because adults do not feed and body mass decreases with age and as a consequence of mating (for males) and egg laying (for females). Having each pair of students run one control and one selection treatment replicate culture will minimize the amount of data collection each student must perform while ensuring adequate replication in a class of 20-30 students. In smaller classes, we have had each student run one control and one bottleneck treatment replicate. Heritability may be a factor in the occurrence of genetic drift since sampling error causing an increase in homozygous genotypes must actually be reflected in a change in phenotype frequency. When heritability is at or near zero, either a population is already homozygous for the gene(s) that underlie a given trait or genotypic differences play a minor role in the observed variation in a trait.
Data analysis—The data from either experiment will be the sex of the adult beetle, its development time or its body mass. The mean development time or body mass should be calculated for each culture along with the minimum, maximum and variance. These statistics should be calculated separately for males and females. In selection experiments, differences between control populations (run simultaneously with the treatment populations) and the treatment populations may be evaluated with two sample t-tests. In genetic drift experiments, the same statistics will be calculated, but statistical tests are not clearly applicable. The expected outcomes of genetic drift are random changes from the control populations, so the average values for all control populations should be compared to each individual bottleneck population. The bottleneck treatment populations could be categorized as having drifted in a positive direction or a negative direction and separate t-tests (or non-parametric equivalent Mann-Whitney U) could be performed to compare those groups with the control populations. Alternatively, the characteristics of adults starting each population (founders) could be compared to the characteristics of the individuals emerging (descendents) from the same populations. Genetic drift may occur in all populations but we expect greater drift (more phenotypic change) in bottleneck populations than in control populations. Comparisons between founder and descendant data could be evaluated using a pairwise test (paired t-test). Additional analyses may be performed by addressing two predictions about population changes expected as a consequence of genetic drift. We expect the variation among bottleneck populations to be greater than that among control population since genetic drift should cause more change in the bottleneck populations than in the controls. This prediction could be tested with Levine’s test for equality of variances. We also predict a pattern in the variation within each independent population. Bottleneck populations are predicted to exhibit less variation (regardless of the direction of drift) than control populations. This prediction could be tested by means of a t-test on the collection of variance values calculated for each population to compare the control and bottleneck populations.
Copyright © by Lawrence S. Blumer and Christopher W. Beck, 2009. All rights reserved. The content of this site may be freely used for non-profit educational purposes, with proper acknowledgement of the source. All other uses are prohibited without prior written permission from the copyright holders.
Data on a genetic bottleneck experiment were collected by Mr. Ben Davids at Morehouse College during the summer 2008. A total of 24 independent trials were conducted that consisted of fourteen control populations, contained 25 randomly chosen males and females, and ten bottleneck populations, contained 5 randomly chosen males and females. Both treatments were conducted in 150 x 25mm Petri dishes containing a single layer of dried mung beans. Among the fourteen control populations, the average body mass of males at emergence was 3.82±0.36 mg (mean±SE) and the average body mass of females at emergence was 4.71±0.19 mg (Figure 1 shows a typical control population frequency distribution).
Among the ten bottleneck populations, six had mean male and female body masses larger than those observed in the control populations, two had average body masses smaller than those of the control populations, and two had mixed results with mean male mass smaller than in the controls but mean female mass larger than in the controls. In the bottleneck experiments, male body mass remained smaller than that of females in every population, as was observed in the control populations. A representative example of the adult mass frequency distribution produced from a bottleneck population in which the body mass for both males and females increased is shown below (Figure 2).
Among the six bottleneck treatment populations in which the mean male and female body masses were greater than those observed in the control populations, male body mass was 4.03±0.21mg (mean±SE) and female body mass was 5.07±0.21mg. The body masses of both males and females from these bottleneck populations were significantly greater than those of males and females from control populations (Mann-Whitney U, Na=14, Nb=6 two-tailed p=0.043 for males and p=0.004 for females).
Figure 1. Frequency distribution of male and female body mass in a control population. These data are for one control population that was started with 25 randomly chosen males and females from which 229 males and 250 females were produced. Mean body mass of males was 3.72±0.59 mg (mean±SD) and mean body mass of females was 4.68±0.69 mg in this population.
Figure 2. Frequency distribution of male and female body mass in a genetic bottleneck treatment population. These data are for one bottleneck population that was started with 5 randomly chosen males and females from which 87 males and 77 females were produced. Mean body mass of males (mean±SD) was 3.97±0.50 mg and mean body mass of females was 5.35±0.60 mg in this population.