Sperm-Female Coevolution in Drosophila

See allHide authors and affiliations

Science  08 Nov 2002:
Vol. 298, Issue 5596, pp. 1230-1233
DOI: 10.1126/science.1076968


Rapid evolution of reproductive traits has been attributed to sexual selection arising from interaction between the sexes. However, little is known about the nature of selection driving the evolution of interacting sex-specific phenotypes. Using populations of Drosophila melanogaster selected for divergent sperm length or female sperm-storage organ length, we experimentally show that male fertilization success is determined by an interaction between sperm and female morphology. In addition, sperm length evolution occurred as a correlated response to selection on the female reproductive tract. Giant sperm tails are the cellular equivalent of the peacock's tail, having evolved because females evolved reproductive tracts that selectively bias paternity in favor of males with longer sperm.

Male reproductive traits appear to evolve more rapidly than other types of character (1). For example, DNA sequence comparisons reveal that male-derived molecules involved in reproduction exhibit a high level of divergence among members of the primate lineage leading to humans (2), between mouse and rat (3), among marine invertebrates (4), and between closely relatedDrosophila species (5). Positive Darwinian selection is the driving force behind this rapid evolution (2–5). Models to explain this pattern include sexually antagonistic coevolution and postcopulatory sexual selection and are largely based on male-female interaction (6–9). In line with this view, a recent study has demonstrated positive Darwinian selection driving rapid evolution of mammalian female reproductive proteins as well (10). Despite the growing number of studies illustrating this pattern at the molecular and genetic level and theory-based explanations (6, 7, 9), few studies have identified the interacting sex-specific phenotypes or have experimentally shown the nature of selection driving their rapid evolutionary divergence (11).

Sperm are the most diverse cell type, evolving so rapidly that their external morphology and ultrastructure provide reliable cues for distinguishing taxa and their phylogenetic relationships (12). Sperm tail length variation is of particular interest, as comparative studies on diverse taxa have found positive relationships between sperm size and the risk of sperm competition (13, 14). A pattern of correlated evolution between sperm length and certain dimensions of the female reproductive tract has been identified in taxa as diverse as birds (15), butterflies (16), and fruit flies (17). Collectively, these studies strongly implicate postcopulatory sexual selection mediated by a component of cryptic female choice referred to as “sperm choice” (18). This process, defined as nonrandom paternity biases resulting from female morphology, physiology, or behavior that occur after coupling, has proved difficult to demonstrate (13,18). We have experimentally explored the adaptive importance of sperm length in D. melanogaster and provide evidence for its evolution by female sperm choice. Drosophila species provide a valuable system for investigating sperm evolution, because variation in sperm length within the genus exceeds variation in the rest of the animal kingdom. At one extreme, D. bifurca have evolved gigantic sperm that are 58.290 ± 0.66 mm long, or about 20 times the total body length of males (19).

We established independent populations of D. melanogasterthat were selected for either increased or decreased sperm length or the length of the females' primary sperm-storage organ, the seminal receptacle (SR) (Fig. 1). The sperm length selection experiment had a single replicate, and the SR length selection experiment had two replicates (each replicate consisting of a long, short, and control population). Next, sperm competition experiments (N = 3 replicates) were conducted in which females from the SR selection lines were each initially mated to a standard “competitor” male (bearing an eye color mutation that allowed offspring paternity to be assigned) and then remated to a male from one of the sperm length selection lines. The proportion of progeny sired by the second male (P2) was then determined. The first replicate had four experimental treatments: short-sperm males and long-sperm males were each competed within short-SR and long-SR females from the first replicate of the SR selection experiment (Fig. 2A). The second replicate had six experimental treatments: the same design as the first replicate with the addition of control-SR females, again using females from the first replicate of the SR selection experiment (Fig. 2B). The third replicate had nine experimental treatments: the same design as the second replicate with the addition of control-sperm males, with females coming from the second replicate of the SR selection experiment (Fig. 2C). Because males and females were all derived from independent selection lines, sperm competition outcomes could not be influenced by unforeseen coevolved traits between the sexes. Finally, within the two SR length selection replicates, sperm length was determined after 37 and 42 generations, respectively, to discern any correlated response in males to selection on the female's reproductive tract.

Figure 1

Responses to bidirectional selection on length of (A and B) the female's seminal receptacle (replicates 1 and 2, respectively) and (C) sperm. Selection line: open squares, short; open circles, control; solid squares, long. Populations from which females (A, B) and males (C) were derived for each of the three replicate sperm competition experiments are indicated by dotted lines. Sperm length of the “competitor” males against which sperm-selection line males were competed is also indicated (solid circles). SR length selection was performed as described (27), except that females were mated multiply and selection was intermittent after the fourth generation. For sperm selection, sperm from 45 males per line per generation were measured as described (35); within each line, progeny from the 15 males with the longest or shortest sperm were selected. Control lines were maintained under identical conditions.

Figure 2

Results of three replicate experiments to determine the proportion of progeny (mean ± SE) sired by males from the sperm-length selection lines within females from the SR-length selection lines. Males: open squares, short-sperm selection line; open circles, control-sperm selection line; solid squares, long-sperm selection line. (A and B) Replicates 1 and 2, respectively, females from first replicate of SR-length selection; (C) replicate 3, females from second replicate of SR-length selection.

Sperm length and SR length both responded to directional selection (Fig. 1), resulting in statistically highly significant increases and decreases in both traits relative to control populations (20). Sperm length exhibited a realized heritability of 0.478 ± 0.286, and SR length exhibited realized heritabilities of 0.366 ± 0.270 and 0.414 ± 0.266 in the two respective replicates. These populations were used in sperm competition experiments to determine the adaptive importance of variation in sperm and SR length and the level of interaction between these sex-specific traits in determining differential male fertilization success. Interpretation of the sperm competition experimental results is based on analysis of paternity for 43,031 offspring from 1044 females (21).

All three replicates of the sperm competition experiment revealed a strong interaction between male and female traits (Fig. 2 and Table 1). First, males from populations with longer sperm had a fertilization success equal to or better than that of males with shorter sperm. Second, and most notably, we consistently found statistically significant SR length × sperm length interaction effects determining male fertilization success. Thus, the performance of males with different sperm lengths was influenced by the SR length of females within which males were competing. In contrast, the female line influence on variation in P2 was inconsistent, being significant in only one of three replicates. The significant interaction between the sexes may explain why studies that considered only female morphology and not sperm length, or vice versa, have found no effect on P2 (22, 23).

Table 1

Analysis of variance and covariance of second male sperm precedence (P2). df, degrees of freedom; MS, type III mean square.

View this table:

Long-sperm males did not perform better or worse than short-sperm males within short-SR females in any replicate (post hoc Scheffe'sF tests: rep. 1, P = 0.25; rep. 2,P = 0.68; rep. 3, P = 0.75). In contrast, long-sperm males sired a significantly greater proportion of progeny than did short-sperm males within control-SR females (rep. 2,P < 0.05; rep. 3, P < 0.01), and this difference was even more pronounced within long-SR females (rep. 1,P < 0.001; rep. 2, P < 0.01; rep. 3,P < 0.001). As the length of a female's SR increased, so did the value of long sperm among potential sires (Fig. 2). Length of the female's SR thus represents the mechanical determinant of postcopulatory female “sperm choice” (13, 18) based on the sperm length of males in D. melanogaster.

A separate experiment showed that results of the P2 experiment were not attributable to variation in the number of sperm transferred by males. Long-sperm males transferred the fewest sperm, although differences were not significant (mean ± SE: long-sperm males, 1428 ± 65, N = 10; short-sperm males, 1584 ± 120,N = 10; control-sperm males, 1546 ± 290,N = 5) as indicated by analysis of variance (ANOVA) (F 2,22 = 0.42, P = 0.66). Further, two-way ANOVA revealed no interaction effect between male and female selection lines on the number of sperm transferred (F 1,16 = 2.29, P = 0.15).

Sperm size may influence their position in the SR. Within the SR of allDrosophila species, the sperm generally appear to be straightened out rather than coiled (17). In D. melanogaster, a small and well-organized group of sperm heads can be observed, 6 hours after copulation, near the proximal end of the SR with their tails extending distally (24). These sperm, by virtue of their location, presumably take precedence over sperm residing more distally within the SR. We speculate that “being the right size” offers some advantage to sperm in occupying this superior position.

The demonstrated relationship between SR length and sperm use (Fig. 2) is predicted to generate linkage disequilibrium between genes for the female preference and the male ornament, resulting in the genetic correlation assumed by good genes and runaway sexual selection models (25). We tested this prediction by examining correlated responses in sperm length within the SR-length selection lines. Each female had ample opportunity to mate with multiple males before producing offspring for each subsequent generation. Thus, maintenance of genetic covariation between SR length and sperm length could arise through female sperm choice.

Sperm length increased significantly in both long-SR selection lines relative to control-SR lines (ANOVA: rep. A,F 2,42 = 32.12, P < 0.0001; rep. B, F 2,42 = 8.99, P < 0.001), but showed no significant change in the short-SR selection lines (Fig. 3) [relationships between sperm length and male body size were not significant (20)]. This result is likely attributable to postcopulatory sexual selection rather than to pleiotropy or genetic linkage. Results of the sperm precedence experiments indicate that short SRs do not discriminate among sperm within the natural distribution of lengths, whereas long SRs do discriminate against shorter sperm (Fig. 2). The observed pattern of evolutionary change in sperm length within the SR selection lines is precisely what one would predict from the results of the sperm competition experiments. No such pattern is predicted as a consequence of pleiotropy or linkage. We suggest that the process of female “sperm choice,” as demonstrated here for D. melanogaster, underlies the marked divergence in sperm length throughout the genus Drosophila and other taxa. If true, then postcopulatory sexual selection clearly can favor sperm quality (e.g., length) at the expense of sperm quantity, even when males have limited resources for gamete production (26).

Figure 3

Correlated response of sperm length (mean ± SE) to selection on length of the female's SR (open squares, short-SR line; open circles, control-SR line; solid squares, long-SR line).N = 15 males per line.

Although we now understand what drives sperm length evolution, we do not know what is driving the evolution of SR length. Nonetheless, this trait offers an exceptionally tractable system for studying the evolution of a female preference and of male-female interactions. The functional relationship between the female preference and the corresponding male ornament is unambiguous, the preference and ornament are both easy to quantify, the macroevolutionary pattern of coevolution between the preference and ornament has been established (17), costs of relative expression of each have been quantified (19, 23, 26), and each is amenable to genetic analysis and artificial selection (27).

Our results are consistent with several models developed to explain the evolution of female mate preferences. Linkage disequilibrium between the female preference and male ornament is consistent with the Fisherian runaway process and “good genes” models (28). Also consistent with good genes models, recent studies have suggested a link between male condition and sperm quality (29), including sperm length (30). Next, interactions between the sexes are rife with conflict in D. melanogaster (31) and the coevolution of sperm and SR length may be sexually antagonistic, as has been suggested for sperm length and sperm-storage tubule length in birds (15). Finally, data reported here refute predictions of two sexual selection models as applied to this system. First, the “direct benefits” model (28) cannot apply, as the long sperm tails are not absorbed by females and have not evolved to serve a post-fertilization function (32). Second, the “sensory exploitation” model (28) is not applicable, as phylogenetic analysis reveals a pattern of correlated evolution between the female preference and male trait (17) rather than a pattern of the male trait evolving in response to a preexisting female bias.

The sperm-female coevolution demonstrated here has important implications for diversification and speciation. Rapid morphological divergence of sperm has been reported for numerous taxa, including primates (33). Such divergence has been shown to drive correlated divergence of important life history traits (19,26). Further, as sperm morphology and sperm usage by females are central to successful reproduction, their divergence will likely contribute to reproductive isolation between populations and the formation of new species (1, 7, 34).

  • * To whom correspondence should be addressed. E-mail: sspitnic{at}


View Abstract

Stay Connected to Science

Navigate This Article