Exercise 6.3

Examining the Evolutionary Significance of Testes Size in Drosophila, or What Happens When Biologists Select for Larger and Smaller Testes?

(This exercise is based on Pitnick, S., and G. T. Miller. 2000. Correlated response in reproductive and life history traits to selection on testis length in Drosophila hydei. Heredity 84: 416–426.)

(Note: The reference above links directly to the article on the journal’s website. To access the full text of the article, you may need to be on your institution’s network [or logged in remotely], so that you can use your institution’s access privileges.)


The genus Drosophila is known for the incredible diversity of sperm size and shape, including giant sperm found in some species. For instance, males of the species D. bifurca have an average sperm length close to 60 mm, twenty times their body length! What is the purpose of such large sperm? What reproductive benefits are obtained by having such large sperm? In contrast, most Drosophila species do not have such large sperm. Why not? Are there costs associated with the production of large sperm? These are among the types of questions that intrigue evolutionary biologists who study sperm size and other reproductive traits.

Variation within natural populations, though abundant, usually consists of very small differences among individuals. The minute size of these differences makes it difficult to test their significance. Evolutionary biologists can use artificial selection as a tool to magnify the differences present in natural populations. An example of this type of study is work done by Scott Pitnick and Gary Miller at Syracuse University. They performed experimental evolution to examine the large sperm and testes of D. hydei. Although the sperm of D. hydei are not as large as those of D. bifurca, D. hydei still has respectably large sperm (23 mm long) and testes (30 mm long).

In earlier work, Pitnick found a nearly perfect correlation between the size of sperm and the size of testes across different species of Drosophila. Because sperm are more difficult to measure than testes, especially when dealing with a large number, Pitnick and Miller decided to select on testis size rather than on sperm size. In other words, they used testis size as a proxy for sperm size.

Pitnick and Miller acquired flies that had been collected from a natural population of D. hydei from California. After these flies acclimated to the laboratory for a few generations, the researchers started the experimental lines. In each generation, for the “high” line, offspring from males with the largest testes were used to begin the next generation. A similar procedure was used for the “low” lines, except that progeny from the males with the smallest testes were used to start each generation.


Figure 1 The response to selection on testis length over successive generations in the high and low lines in each of the replicates. Each symbol represents the population mean for a particular line at a given generation. The y-axis displays the testis length in millimeters. On the x-axis is the cumulative selection differential, the sum of selection differentials from each generation.


Question 1. Refer to Figure 1 above. Did the testis length respond when selection was applied in the up direction?


Question 2. Did the testis length respond when selection was applied in the down direction?


Question 3. Was the response equal in the two directions? By how much did each line respond?


Question 4. By about how much are the testes of males from high lines larger than testes of males from low lines?


Question 5. This is a more difficult question. From Equation (6.1) on page 145, we see that the heritability can be estimated by dividing the evolutionary change in a single generation, $\Delta \overline{Z}$, by the selection differential, S. We can also estimate the heritability using the evolutionary change over several generations. To do that, in place of the selection differential for a single generation, S, we use the “cumulative selection differential.” This is simply the sum of all the selection differentials in each of the generations. In the experiment, the cumulative selection differential was about 25mm for the high lines and –18 mm for the low lines. Pitnick and Miller selected only on males, and not both males and females. To correct for the fact that the investigators selected on only one sex, the heritability is actually twice the value given by Equation (6.1). What are the heritabilities for the high and low lines from replicate A?


Figure 2  The average sperm size for the high (clear) and low (shaded) selection lines for replicate A and B, respectively on the left and the right.


Question 6. As previously noted, one important use of selection lines is to examine correlated responses to selection. That is, have traits other than the trait being selected on changed during the selection, and if so, in what ways? Given the striking correlation of sperm size and testis length across different Drosophila species, one might expect sperm size to evolve as a indirect response to testis-size selection. According to Figure 2 above, did sperm length evolve as a indirect response to selection on testis length? If so, was the correlation of the change in sperm length and testis length positive or negative? Explain.

Question 7. How much difference exists in the sperm length between the high and the low lines? Express this figure as a percentage of the sperm length of the lines with the lower average sperm length.


Figure 3

Question 8. Refer to Figure 3 above. Was there a correlated response of thorax length that resulted from selection on body size? If so, was the correlation of the responses positive or negative?


Question 9. What is the difference in the thorax size between the high and the low lines in terms of a percentage of the thorax size in the low lines in both replicate A and replicate B? How does the magnitude of the change in thorax length compare with that of the change in testes size (see Question 4)? Elaborate.


Figure 4 The mean value of development time (in hours) on the y-axis for high line males (white bars), low line males (pixilated bars), high line females (grey bars), and low line females (black bars). Replicate A is on the left, and replicate B is on the right.


Question 10. Because the production of large testes requires the allocation of more resources to the testis, one might expect a correlated response in development time when testis size is changed. Refer to Figure 4 above. Does development time vary between the high and low lines, and if so, in what direction?


Question 11. Which sex varies the most in development time between the high and low lines?


Table 1  The percentage of males of various ages that have mature sperm.


Question 12. Refer to Table 1 above. Do the high and low testis size lines differ with respect to the time to develop mature sperm? If so, how?


Question 13. Based on what you have learned above, is there evidence for cost associated with large testes? Explain.


Question 14. Based upon the evidence with which you have been provided, can you infer that sexual selection is operating on testis length? If yes, state the data that supports the action of sexual selection. If no, what type of evidence would support the action of sexual selection on this trait?