- Email: [email protected]
The risk of sperm competition and the evolution of sperm heteromorphism
ANIMAL BEHAVIOUR, 1998, 56, 1497–1507 Article No. ar980930
The risk of sperm competition and the evolution of sperm heteromorphism RHONDA R. SNOOK
Department of Zoology, Arizona State University, Tempe (Received 3 February 1998; initial acceptance 27 February 1998; final acceptance 30 March 1998; MS. number: A8025R)
ABSTRACT Members of the Drosophila obscura species group exhibit sperm heteromorphism in which males simultaneously produce two different morphologies of sperm, short and long. Short sperm represent at least 50% of the ejaculate in several species but do not function in fertilization and thus, the evolutionary significance of this phenomenon is unknown. Selective pressures associated with the risk of sperm competition have been suggested to explain the maintenance of sperm heteromorphism. Here I test: (1) whether males alter their behaviour or ejaculate characteristics in response to the risk of sperm competition, and (2) whether short, nonfertilizing sperm serve as ‘cheap filler’ in the female reproductive tract, thereby delaying female receptivity to subsequent matings. Males did not alter tactics based on the risk of sperm competition; copulation durations of males mated to nonvirgin females were unexpectedly shorter than when both sexes were virgin and, contrary to sperm competition predictions, males did not alter the ratio of fertilizing:nonfertilizing sperm relative to female mating status or female age. Short sperm had a small, if any, role in influencing female remating behaviour. The number and proportion of short sperm present in sperm storage organs did not differ between nonvirgin receptive and nonvirgin nonreceptive females. The proportion of short sperm present in the ventral receptacle, however, was lower in receptive than nonreceptive females. Remating behaviour was strongly linked to oviposition patterns. Nonvirgin receptive females oviposited more eggs than nonreceptive females, and female remating interval was positively related to both the number of eggs and progeny produced. Although, oviposition parameters are related to fertilization and thus, the use of long sperm, nonvirgin receptive and nonreceptive females did not predictably differ in the number or proportion of long sperm in storage. These results suggest oviposition per se is probably more important than sperm in determining female receptivity and that sperm heteromorphism may play a marginal, if any, role in affecting female remating.
1994; Snook 1995; Bressac & Hauschteck-Jungen 1996). Both sperm types are nucleated, have the same DNA content (Pasini et al. 1996) and are motile, suggesting both types of sperm can fertilize eggs. Direct measurements of sperm in eggs, however, found that only long sperm fertilize eggs in several obscura group species (D. pseudoobscura, Snook et al. 1994; D. persimilis, D. affinis, D. miranda, D. subobscura and D. athabasca, Snook & Karr 1998; and probably D. obscura, Snook 1997). The evolutionary significance of sperm heteromorphism is unknown. Interpreting the functional significance of this phenomenon in the obscura group is confounded by the fact that all obscura species so far examined produce two discrete types of sperm (Beatty & Sidhu 1970; Sanger & Miller 1973; Joly & Lachaise 1994; Snook 1997), whereas no other Drosophila species exhibits
Males of many species produce multiple sperm types within each ejaculate, a phenomenon known as sperm heteromorphism (Sivinski 1980; Silberglied et al. 1984). For example, lepidopterans produce both small, apyrene (anucleated) sperm, which do not function in fertilization, and larger eupyrene (enucleate) fertilizing sperm (Silberglied et al. 1984). Males of the Drosophila obscura group produce two types of mature sperm, ‘short’ sperm which have short heads and tails and ‘long’ sperm which have long heads and tails (Beatty & Sidhu 1970; Sanger & Miller 1973; Joly & Lachaise 1994; Snook 1997). Short sperm represent at least 50% of the ejaculate (Snook et al. Correspondence and present address: R. R. Snook, Department of Biological Sciences, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154-4004, U.S.A. (email: [email protected]). 0003–3472/98/121497+11 $30.00/0
1998 The Association for the Study of Animal Behaviour
1497
1998 The Association for the Study of Animal Behaviour
1498 ANIMAL BEHAVIOUR, 56, 6
this phenomenon. The production of multiple sperm types in any particular species, therefore, may reflect common ancestry rather than adaptive function (although similarity by descent may be selectively maintained; Brooks & McLennan 1991; Harvey & Pagel 1991). Recent phylogenetic autoregression analyses, however, revealed that 22% of the interspecific variation in the length of long sperm is related to phylogeny, whereas the length of short sperm is not significantly correlated with phylogeny, indicating different selection pressures have acted on the two sperm lengths (Snook 1997). Long sperm have evolved in response to fertilization demands, whereas short sperm have been decoupled from these requirements and may be an adaptation to currently unidentified selection pressures (Snook 1997; Snook & Karr 1998). Several selection pressures have been suggested to explain the maintenance of nonfertilizing gametes, namely that nonfertilizing sperm serve as nutrient contributions (Gage & Cook 1994; Cook & Gage 1995; Snook & Markow 1996) and nonfertilizing sperm function in sperm competition (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995; Snook & Karr 1998), where more than one male’s ejaculate competes within the female reproductive tract for fertilization opportunities (Parker 1970). In both lepidopterans and Drosophila, nonfertilizing sperm do not function as nutrient contributions (Cook & Gage 1995; Snook & Markow 1996). Nonfertilizing sperm in lepidopterans may function in sperm competition (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995), but sperm competition hypotheses have not been tested in the obscura group. Both male behaviour (i.e. copulation duration) and ejaculate characteristics may be influenced by the perceived risk of sperm competition (Smith 1984; Birkhead & Møller 1992). As copulation duration increases, paternity tends to increase. Males of certain species prolong copulation with previously mated females, allowing males to increase paternity by: (1) delivering more sperm (Dickinson 1986), (2) mate guarding (Parker 1970; McLain 1989), or (3) volumetrically removing more of a rival male’s sperm from the female sperm storage organs (Siva-Jothy 1987). Ejaculate characteristics, such as the number of sperm, or in sperm heteromorphic systems, the proportion of sperm types, can also change with the risk of sperm competition. Males may alter the number and/or type of sperm transferred to females in response to: (1) female mating status (Smith 1984; Birkhead & Møller 1992; Wedell 1992), (2) presence of rival males (Gage 1991; Simmons et al. 1993; Gage & Barnard 1996), or (3) female age (Cook & Gage 1995; Gage & Barnard 1996). For example, in sperm heteromorphic lepidopterans, males transfer more fertilizing sperm to previously mated females and transfer fewer nonfertilizing sperm to older females (Cook & Gage 1995). These results indicate that males increase their reproductive success by transferring larger numbers of fertilizing sperm to compete with a prior male’s sperm and suggest that, because younger females are likely to have higher future fecundity, males transfer more nonfertilizing sperm to delay female
remating behaviour and prolong the use of their sperm (Cook & Gage 1995). The effect of sperm competition on ejaculate and behavioural traits in sperm heteromorphic obscura group species is not known. Nonfertile sperm morphs may play a specific role in sperm competition by either functioning to delay female remating by serving as ‘cheap filler’ within female sperm storage organs (SSOs) or displacing/incapacitating rival sperm already in storage (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995). Several aspects of Drosophila biology suggest that short, nonfertilizing sperm could function in sperm competition by serving as cheap filler to delay female remating. First, in all obscura group species examined, females multiply mate so that sperm from more than one male competes to fertilize eggs (Cobbs 1977; Pruzan-Hotchkiss et al. 1981). Second, in many Drosophila species, as with lepidopterans, long-term inhibition of female remating is linked to sperm load (Manning 1962; Drummond 1984; Gromko et al. 1984; Schwartz & Boake 1992; Gromko & Markow 1993; Pitnick & Markow 1994a; He et al. 1995), either through the positive correlations between female remating behaviour and progeny production, an indicator of sperm use (Trevitt et al. 1988; Gromko & Markow 1993), or between the number of sperm received per copulation and female remating interval (Pitnick & Markow 1994a). Third, short sperm are probably relatively cheaper to produce than long sperm as Drosophila males that produce longer sperm mature later than males that produce shorter sperm (Pitnick et al. 1995). Here I test the hypothesis that sperm heteromorphism is an adaptation to sperm competition by assessing whether: (1) males alter their copulation duration or ejaculate characteristics in response to the risk of sperm competition, and (2) short sperm serve as cheap filler delaying female receptivity (Silberglied et al. 1984). METHODS
Fly Maintenance T. Markow collected D. pseudoobscura in February 1991 from fallen citrus in Tempe, Arizona, D. affinis (14012–0141.2) was obtained from the National Drosophila Species Resource Center, Bowling Green, Ohio, and D. persimilis was kindly provided by A. T. Beckenbach (Simon Fraser University, Burnaby, British Columbia). All cultures were maintained on standard cornmeal–agar–molasses food with yeast and kept at room temperature, 22–25C, and an approximate 12:12 h light:dark cycle. I collected virgin flies, separated males and females upon eclosion by aspiration, and stored flies 10 per 10-dram yeasted food vials. All flies used were 5 days old, unless otherwise noted, and reproductively mature (Snook 1995).
Sperm Counts I determined the ratio or number of sperm types by dissecting sperm from either male seminal vesicles or
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1499
female reproductive tracts in phosphate-buffered saline (PBS) on a microscope slide, fixing the sperm, and then staining the sperm with Hoescht’s 33258, a DNA-specific dye, as described previously (Snook et al. 1994). I then counted sperm on an epifluorescent microscope. Males produce and transfer to females too many sperm to enumerate completely so I counted 500 sperm per sample to get proportion data for seminal vesicles and uterus sperm counts. Sperm counts in SSOs were based on total enumeration because fewer sperm are found in these organs (Snook et al. 1994).
Copulation Duration, Female Mating Status, Female Age and Sperm Transfer Patterns I performed interrupted copulations on D. pseudoobscura to determine whether the two sperm types were transferred at different times during copulation. This was assessed as a preliminary examination of the role sperm length may play in the timing of the entrance into SSOs and the competitive ability of different sperm lengths, two parameters previously invoked as determinants of the outcome of sperm competition (Gomendio & Roldan 1991; Briskie & Montgomerie 1992; Pitnick & Markow 1994a). I paired two virgin males and one virgin female and interrupted copulations at 30-s intervals up to 3 min by either gently aspirating the mating pair or tapping the vial. I dissected females immediately after copulation was interrupted and determined the proportion of sperm types in the uterus. I examined the influence of female mating status on copulation durations in three species, D. pseudoobscura, D. persimilis and D. affinis. Males mated to nonvirgin females are predicted to have longer copulation durations than when mated to virgin females. I paired two virgin males and one virgin female for mating and subsequently exposed mated females to virgin males for 2 h in the morning and afternoon on each of 5 consecutive days (periodic interaction design; Pyle & Gromko 1978). I recorded durations of all copulations. Additionally, I examined the influence of male mating status on copulation duration in two species, D. pseudoobscura and D. persimilis. For this experiment, I paired two virgin females and one virgin male. I removed the females when copulation had ended and replaced them with two new virgin females until the next copulation occurred. I recorded the durations of all copulations. To determine whether female mating status influenced the proportion of sperm types males transferred, I paired virgin D. pseudoobscura males to nonvirgin females that had mated 48 h earlier. Immediately following the females’ second copulation, I dissected females and determined the proportion of sperm the second male transferred (sperm storage does not begin until ca. 2 h after copulation). I then compared these data with the proportion of sperm transferred by virgin males to virgin females (data from Snook et al. 1994). To determine whether female age influenced the proportion of sperm types transferred to females, I paired virgin males and virgin females that were either 5 or 15 days old. Immediately following copulation, I dissected
the females and determined the proportion of sperm transferred.
The Cost of Sperm Production For short sperm to function as cheap filler, they should be less costly to produce than long sperm. One measurement of cost is the rate at which males produce each sperm type. I paired one set of D. pseudoobscura males to virgin females once every day for 15 days (Daily mating treatment), and paired another set of males to virgin females once every other day (Alternate mating treatment). I discarded data for individual males if the males failed to remate anytime during the experiment. Finally, I maintained another set of virgin males and did not pair them with females for 15 days (Virgin 15-day treatment). On the 16th day of the experiment, I dissected all males and processed the seminal vesicles for sperm counting. I maintained all females that mated with males during Daily or Alternate mating treatments in individual vials, and transferred the females to new food vials every other day for 5 days. I retained all vials for subsequent progeny counts and compared the number of progeny produced between females mated to males of the two mating groups.
Sperm Load, Female Receptivity and the Cheap Filler Hypothesis To determine whether short sperm increase female remating latency, I tested the prediction that nonvirgin receptive females should have fewer short sperm (less filler) than nonvirgin nonreceptive females. I paired two virgin D. pseudoobscura males with a virgin female in a yeasted food vial. Upon mating, I removed the nonmating male from the vial by aspiration and, following the completion of copulation, removed the mating male. Thirty-six hours later, I paired two virgin males with a previously mated female in a yeasted food vial for a maximum of 2 h. Thirty-six hours was chosen because the average D. pseudoobscura female remating interval is 2 days (48 h; see below). A female was considered receptive to remating if copulation occurred (the male mounted and inserted his genitalia into the female). When this copulation occurred, I immediately aspirated the pair apart prior to sperm transfer, removed the male from the vial, and immediately dissected the female for sperm counts. If females did not indicate a willingness to remate by the end of the 2-h receptivity test, they were conidered nonreceptive. In these cases, I removed males from the vials, and immediately dissected the nonreceptive females for sperm counts. I also counted the number of eggs oviposited in vials that held females during the time between the first mating and the receptivity test to determine sperm use prior to remating. I discarded females that died on or before day 5 of the experiment. I used one-tailed t tests to determine whether nonvirgin nonreceptive females and nonvirgin receptive females differed in the sperm types stored because the prediction specifically states that the mean number or proportion of sperm types stored by nonvirgin
1500 ANIMAL BEHAVIOUR, 56, 6
nonreceptive females should be greater than receptive females (ìnonreceptive >ìreceptive).
1.0 1/8
Sperm Use and Female Remating Patterns
RESULTS
Copulation Durations, Female Mating Status, Female Age and Sperm Transfer Patterns Virgin males did not begin to transfer sperm to virgin females until approximately 1.5 min into copulation (Fig. 1). Both sperm types were transferred to females at the start of sperm transfer (Fig. 1). Furthermore, the proportion of short sperm transferred by males did not differ, either during any 30-s interruption of copulation or from complete copulations (ANOVA: F4,30 =0.264, P=0.9; Fig. 1). Female and male mating status influenced copulation duration. In all three species examined, the copulation durations of males mated to previously mated females were significantly less than that of males mated to virgin females (Table 1). Similarly, copulation durations of females mated to previously mated males were significantly less than that of females mated to virgin males in D. pseudoobscura and D. persimilis (D. affinis was not tested; Table 1). Female mating status did not influence the ratio of sperm types males transferred to females. The proportion of short sperm transferred by virgin males to virgin females (0.4790.023, N=11) did not differ from the proportion transferred to females that had mated 48 h earlier (0.5080.029, N=13: t22 =0.761, P=0.45). Female age did not influence the ratio of sperm types transferred to females. The proportion of short sperm
0.8 Proportion of short sperm
To estimate the relationship between female remating and sperm use, I examined the correlations between egg and progeny production and remating interval in D. pseudoobscura, D. persimilis and D. affinis. I allowed females the opportunity to mate for 2 h in the morning and afternoon for 5 consecutive days. I transferred females to new food vials containing dried yeast paste once a day and/or after remating for the first 15 days, and then every other day until death. I retained all vials for subsequent adult progeny counts. I discarded data for individual females if the females did not remate or failed to produce progeny. I transferred singly mated control females to new food vials and treated them as described for multiply mated females. I performed two to six replicates of each treatment, either multiply mated or singly mated females, in each species. I tested replicates for homogeneity using independent t tests or ANOVA. If the replicates were homogeneous then I pooled data. I subsequently performed a linear regression analysis to describe the relationship between female remating and the production of eggs and progeny. I also compared the total number of eggs oviposited (fecundity) and progeny produced (productivity) by singly or multiply mated females. I performed all statistical analyses using SYSTAT (Wilkinson 1990).
0.6
4/9
12/14 2/9
0.4
11/11
6/11
0.2
0/6
0
0.5
1.0
1.5 2.0 2.5 3.0 Copulation duration (min)
6.0*
Figure 1. The proportion (±SE) of short sperm found in the uterus of females following the interruption of copulation at 30-s intervals up to 3 min and a full copulation. *Data from Snook et al. (1994). Numbers in bars indicate the number of males that transferred sperm (numerator) in the number of pairs tested (denominator).
transferred by virgin males to 5-day-old virgin females (0.4730.013, N=20) did not differ from the proportion transferred to 15-day-old virgin females (0.4770.017, N=19: t37 =0.838, P=0.84).
The Cost of Sperm Production Male mating behaviour influenced sperm production patterns in that males given either daily or alternate (every other day) opportunities to mate produced a significantly higher proportion of short sperm than virgin males of any age examined (F3,47 =8.380, P=0.0001; Fig. 2). The proportion of short sperm produced by males that mated daily did not differ from that produced by males that mated every other day (Fig. 2). The number of progeny produced by females that mated to either daily mated males or alternately mated males did not differ (t20 =1.794, P=0.09; Fig. 2).
Sperm Load, Female Receptivity and the Cheap Filler Hypothesis Nonvirgin receptive D. pseudoobscura females oviposited more eggs, suggesting they had used more sperm, than nonvirgin nonreceptive females (t103 = 2.336, P=0.01; Fig. 3). Moreover, 50 of 65 nonreceptive females had an egg in the uterus upon dissection, whereas 63 of 66 receptive females did not (G=81.18, P<0.001). Despite these apparent differences in sperm use, nonvirgin receptive and nonreceptive females had the same total number of sperm in storage (irrespective of sperm type and sperm
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1501
Table 1. The copulation duration of virgin females mated to virgin males, and those same females when they remated, and the copulation duration of virgin males mated to virgin females, and those same males when they remated, and the sample size (N), paired t, and P for each comparison of three obscura group species Copulation duration (min)
Copulation duration (min)
Species
Virgin /
Remating /
N
Paired t
P
Virgin ?
Remating ?
N
Paired t
P
D. pseudoobscura D. persimilis D. affinis
6.2±0.2 7.9±0.3 1.4±0.1
3.7±0.2 5.2±0.3 1.1±0.1
52 45 24
8.597 5.374 2.598
<0.001 <0.001 0.02
6.4±0.2 5.4±0.3
4.4±0.4 4.6±0.4
43 22
7.141 2.047
<0.001 0.05
storage location; t46 =1.191, P>0.1; Fig. 3), suggesting total sperm load was not related to female remating behaviour. No significant differences in the number of either short or long sperm were observed between nonvirgin receptive and nonreceptive females in any comparison (Fig. 4). The proportion of either short or long sperm in the spermathecae and all SSOs also did not differ between females (Fig. 4). The proportion of long sperm in the ventral receptacle of nonreceptive females, however, was significantly less than that of receptive females, contrary to predictions that sperm load influences female remating behaviour. The proportion of short sperm in the ventral receptacle of nonreceptive females was significantly greater than that of receptive females, supporting the
0.75 B B
hypothesis that short sperm serve as cheap filler (Fig. 4).
Sperm Use and Female Remating Patterns Female D. pseudoobscura usually mated three times during the 5-day mating portion of the experiment (median number of mates=3, total number remating=50/60), and the median remating interval from the initial mating to the first remating was 2.0 days. Females oviposited 69.4 (4.5) eggs and produced 52.7 (3.7) progeny between the initial mating and the first remating. There was a significant positive correlation between remating interval and the number of either eggs or progeny produced (Fig. 5a). Of the 50 females that initially remated, 39 remated again. In this case, the median remating interval was 1.5 days between the first remating and the second remating. Between the first and second remating, females oviposited 47.0 (4.6) eggs and produced 33.9 (3.7) progeny. There was also a significant positive correlation
11
0.50
A
1300
13 16
1200
Nonreceptive Receptive
1000
0.25
0.00
24 24
1100
Number
Proportion of short sperm
11
A
Virgin 5 days
Virgin 15 days
Daily
Alternate
900 800 700 200 P < 0.05
Treatment Figure 2. The proportion (±SE) of short sperm found in the seminal vesicles of 5- and 15-day-old virgin males, and in males given the opportunity to either mate every day for 15 days (Daily), or every other day for 15 days (Alternate). Different letters denote significant differences between groups. Females in the Daily mating treatment produced 601.5±31 progeny, and females in the Alternate mating treatment produced 683±26 progeny. Numbers in bars indicate the sample size.
100
0
54
51
Eggs
Total sperm
Figure 3. The mean (±SE) number of eggs and total number of sperm (short+long) found in all SSOs of nonreceptive and receptive females. Numbers above bars are sample sizes. See text for statistics.
1502 ANIMAL BEHAVIOUR, 56, 6
200
Nonreceptive Receptive
t 47 = –1.351
(a) 175
0.25 t 47 = 1.677, P = 0.05
150 0.20
125 100
0.15
75
0.10 t 47 = 0.887
50
0.05
25 24
25
0
1000 900
t 47 = 0.979
(b)
0.20 t 47 = 0.816
700
Proportion of short sperm
Number of sperm
800
600 500 400 300
t 47 = 1.289 24
200
25
100 0
0.15
0.10
0.05
0
1200 1100
(c)
t 46 = 0.966
1000
0.20
900
t 46 = 0.981
800 0.15
700 600 500
0.10
400 300 200
t 46 = 1.485 24
0.05 24
100 0
Long
Short
0
Sperm type Figure 4. The number of short and long sperm and the proportion of short sperm found in the (a) ventral receptacle, (b) spermathecae and (c) all SSOs, in nonvirgin nonreceptive and receptive females. Numbers above bars are sample sizes for both sperm types. All comparisons, except the proportion of short sperm in the ventral receptacle, were NS at P>0.05.
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1503
140 120
140 (a) N = 50
120
100
100
80
80
60
60
40
40
20
20
0
0
(b) N = 39
120 (c) N = 50 Number of eggs and progeny
100
80
60
40
20
0 100
100 (d) N = 26
(e) N = 13
80
80
60
60
40
40
20
20
0
1
2
3
4 0 Remating interval (days)
1
2
3
4
Figure 5. The relationship between the number of eggs oviposited and progeny produced either between the initial mating and the first remating (a, c, d), or between the first and second remating (b, e), and the remating interval for D. pseudoobscura (a, b), D. persimilis (c), and D. affinis (d, e). Remating interval was calculated as the number of days between one mating and the next. (a) Eggs: R2 =0.41, P<0.001; progeny: R2 =0.39, P<0.001, (b) eggs: R2 =0.26, P=0.001; progeny: R2 =0.18, P<0.01, (c) eggs: R2 =0.18, P<0.01; progeny: R2 =0.16, P<0.01, (d) eggs: R2 =0.32, P<0.01; progeny: R2 =0.312, P<0.05, (e) eggs: R2 =0.53, P<0.01; progeny: R2 =0.17, P>0.05. x: eggs; _: progeny.
1504 ANIMAL BEHAVIOUR, 56, 6
in fecundity and productivity between the first and second remating (Fig. 5b). There was no significant difference in the average time to remating between the initial–first remating and the first–second rematings for D. pseudoobscura females (paired t38 =1.188, P=0.2). Females produced fewer progeny between the first and second remating interval compared with the number produced between the initial and first remating interval (paired t38 =2.118, P=0.04), but fecundity did not differ between remating bouts (paired t38 =1.904, P=0.06). Female D. persimilis tended to remate only once during the remating test (median number of mates=2, total number of females remating=50/60, median remating interval=2.375 days). Irrespective of remating intervals, D. persimilis females oviposited 55.8 (3.2) eggs and produced 30.7 (3.0) progeny between the initial mating and the first remating. This approximately 50% reduction in the number of eggs oviposited compared with progeny produced may be a function of unsuccessful fertilizations (recent analyses have found that half the eggs oviposited are not fertilized; Snook & Karr 1998). There was a significant positive correlation between remating interval and both the number of eggs and progeny produced (Fig. 5c). Of the 50 females that initially remated, only five remated again, therefore I did not analyse the association between the first and second remating interval and fecundity and productivity. Half the D. affinis females remated once, and the other half remated twice (median number of mates=2, total number of females remating=28/30, median remating interval=1.5 days). Females oviposited 47.5 (3.2) eggs and produced 22.8 (2.8) progeny between the initial mating and the first remating. As with the other two species, there was a significant positive correlation between the initial remating interval and the number of eggs and progeny produced (Fig. 5d). Of the 13 females that remated twice, the median remating interval was 2.0 days after the initial remating. These females oviposited 47.8 (6.1) eggs and produced 17.3 (3.2) progeny between the first and second remating. Again, there was a significant positive correlation between remating interval and egg and progeny production (Fig. 5e). Female D. affinis took significantly longer to remate the second time than they did between their initial mating and first remating (paired t12 =2.843, P=0.01). Fecundity and productivity, however, did not differ between the initial mating and the first remating and the first and second remating intervals (eggs: paired t12 =0.421, P=0.7; progeny: paired t12 =0.76, P=0.5).
DISCUSSION
Alteration of Behavioural and Ejaculate Characteristics Ejaculate characteristics and male behaviour in many species, including those that are sperm heteromorphic, are influenced by the perceived risk of sperm competition (Smith 1984; Bellis et al. 1990; Gage 1991; Birkhead &
Møller 1992; Parker & Simmons 1994; Cook & Gage 1995). However, obscura group males do not alter ejaculate (proportion of short and long sperm) or behavioural (copulation duration) characteristics in response to a perceived risk of sperm competition. Drosophila pseudoobscura males transfer both sperm types simultaneously during copulation to virgin females and both sperm types enter SSOs at approximately the same time (Snook et al. 1994; Snook 1995). These results indicate that sperm length per se does not influence the timing of sperm entrance into SSOs and suggests that sperm length may not influence sperm competitive ability as has been proposed for sperm monomorphic systems (Gomendio & Roldan 1991; Briskie & Montgomerie 1992). Additionally, these results suggest that short sperm are not ‘front runners’, which serve to actively displace a previous male’s sperm (Silberglied et al. 1984), because both sperm types are transferred and stored simultaneously. Males also do not alter the proportion of short and long sperm based on female mating status. In lepidopterans, males also do not increase the number of nonfertilizing sperm based on female mating status, but do increase the absolute number of fertilizing sperm (Cook & Gage 1995). Although I did not measure the absolute number of short and long sperm transferred to females of different mating status, I found no proportional changes. Male D. pseudoobscura transfer approximately 50% of each sperm type to females; therefore, any potential differences in the absolute number of sperm transferred would have been equal for both types of sperm. In the Indian meal moth, Plodia interpunctella, males transfer more nonfertilizing sperm to young females compared with older females (Cook & Gage 1995). This difference may ultimately be a result of younger females having a longer potential life span, thus a higher reproductive value and increased risk of encountering other males, which males attempt to counter by transferring more ‘cheap filler’ to younger females (Cook & Gage 1995). However, in D. pseudoobscura, female age did not influence the proportion of each sperm type transferred by females. Five-day-old virgin females received the same proportion of short and long sperm as did 15-day-old virgin females. Predictions about the causal factors influencing variation in copulation duration have been described for some species. These factors are complex and can depend on the form of sperm precedence (Yasui 1994), female mating status and oviposition patterns (Siva-Jothy 1987), size of males (Parker & Simmons 1994), and age of males (Siva-Jothy 1987). In general, however, longer copulations lead to a higher reproductive success for males. In the three obscura group species examined, males that mated with nonvirgin females and thus, experienced sperm competition, copulated for an unexpected shorter duration than males that mated with virgin females. This supports a previous study examining copulation duration in D. affinis (Bressac et al. 1991a). In D. pseudoobscura and D. persimilis, male mating status also influenced copulation duration. Mated males copulated for a shorter duration with virgin females than did virgin males. Thus, under the scenarios tested, I found no evidence that
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1505
males respond predictably to sperm competition risks through ejaculate or behavioural changes.
The Cheap Filler Hypothesis Studies on Drosophila that produce giant sperm have found that long sperm are costly to produce (Pitnick & Markow 1994b; Pitnick et al. 1995; Pitnick 1996), thus intuitively, within the obscura group, short sperm are likely to be less costly to produce than long sperm. Here I found that frequent mating by D. pseudoobscura males results in males producing a lower proportion of long sperm compared with virgin males of any age. These results suggest that it takes longer to produce long sperm, and thus, reflects a two-fold cost to producing long sperm: increased time to produce, and when frequently mating, a decrease in the number of long sperm available for fertilization. Short sperm are cheaper to produce than long sperm, but do they function as filler? To function as filler, short sperm must influence female remating behaviour. Previous studies linked female remating behaviour to sperm load by examining the relationship between progeny production and female remating (e.g., Gromko et al. 1984; Schwartz & Boake 1992; Gromko & Markow 1993; He et al. 1995). I found a significant positive correlation between remating intervals and both fecundity and productivity in all three obscura species examined. Female D. melanogaster that oviposit more eggs or produce more progeny tend to remate before females that oviposit or produce fewer eggs or progeny, and this evidence has been used to link sperm load and female remating behaviour (Trevitt et al. 1988). Moreover, field-caught D. melanogaster and D. simulans females that remate produce fewer progeny than females that do not remate, suggesting that females wait to remate until ‘a substantial number of sperm’ from a previous mating have been used (Gromko & Markow 1993). Here I found similar results: receptive females oviposited more eggs, and thus used more sperm, than nonreceptive females. If remating interval is determined by sperm load, however, then the number of eggs oviposited and/or progeny produced between the initial mating and the first remating and the first and second remating intervals should be the same, assuming males transfer and/or females store similar numbers of sperm after each mating. (I did not directly test this assumption, however, males do not alter the proportion of sperm types transferred to females.) Contrary to this prediction, the productivity between the initial and first remating was significantly greater than the productivity between the first and second rematings for D. pseudoobscura. Additionally, even if a second male transfers more sperm to a mated female to better compete with a prior male’s sperm, productivity between the first and second remating should be higher, rather than lower as was found. In D. affinis, females oviposited the same number of eggs and produced the same number of progeny between remating episodes despite the second remating interval being significantly longer than the first, also indicating sperm use had little affect on remating interval. Other studies found discrepancies between the
number of females remating based on oviposition patterns and the association with sperm load. For example, Pruzan-Hotchkiss et al. (1981) found that there was no difference in the proportion of D. pseudoobscura females remating at 3 and 6 days after their initial mating despite a three-fold difference in stored sperm use. If sperm load determined remating, then a higher proportion of females would be expected to remate at 6 days compared with 3 days. These combined data suggest sperm load does not effect female remating in D. pseudoobscura. The confounding element in these results, however, is that indirect fecundity and productivity data only address the influence of long sperm on remating interval because only long sperm are used for fertilization (Snook et al. 1994; Snook & Karr 1998). In the present study, there was no difference between the number and proportion of long sperm found in SSOs of nonvirgin receptive and nonreceptive females, which supports the notion that long sperm do not mediate female remating interval. In lepidopterans, Cook & Gage (1995) found males transfer more nonfertilizing sperm to younger females that are more likely to remate compared with older females. Additionally, the number of nonfertilizing sperm in the armyworm, Pseudaletia separata, decreases within 3 days after copulation, with most disappearing within the first 24 h, and this disappearance correlates with female remating behaviour (He et al. 1995). Similarly, in all three obscura group species examined here, the number of short sperm in SSOs was severely reduced 48 h after copulation (Snook et al. 1994; Snook 1995). The significance of this reduction is unknown; however, the reduction does not reflect their use in fertilization (Snook et al. 1994; Snook & Karr 1998) or the incorporation of nutrient donations (Snook & Markow 1996), nor does it appear to influence female remating behaviour. In the present study, nonvirgin receptive and nonreceptive females did not differ in the number of short sperm stored in any SSO. The proportion of short sperm stored in the ventral receptacle of nonreceptive females was significantly greater than in receptive females, but the proportion of short sperm in the spermathecae and the SSOs combined were not significant. Most D. pseudoobscura sperm are found in the spermathecae (Snook et al. 1994) and sperm from all SSOs are used equally (Patterson 1954). Given this, if short sperm influence female remating behaviour, then nonvirgin nonreceptive females would be expected to have more filler in the spermathecae or both SSO types, contrary to the proportional difference found only in the ventral receptacle.
Sperm Motility, Oviposition and Female Remating Behaviour In lepidopterans, the presence of sperm in the spermatheca delays female remating behaviour (Drummond 1984), and this behavioural modification is thought to be mediated by sperm motility (Thiobold 1979). Nonfertilizing sperm are initially more motile than eupyrene sperm so they enter SSOs first, subsequently generating motility
1506 ANIMAL BEHAVIOUR, 56, 6
and delaying female remating (Silberglied et al. 1984; Cook & Gage 1995; He et al. 1995). Additionally, Gage (1994) found no relationship between the risk of sperm competition and nonfertilizing sperm length, but this length did correlate with male body size. Gage (1994) suggested that if body size correlates with female reproductive tract morphology, then nonfertilizing sperm may function as cheap filler in SSOs through motility differences between sperm types. Female D. melanogaster receptivity is proximately controlled by the central nervous system (Tompkins & Hall 1983). A mechanical stimulation of the central nervous system through sperm motility has been suggested to influence female remating (Gromko et al. 1984). Sperm motility may be detected by the innervation of the ventral receptacle (Miller 1950) and so as the number of sperm in storage decreases below some threshold, the detection of sperm motility would decrease, causing females to remate (Gromko et al. 1984). Drosophila sperm exhibit two kinetic parameters, minor (beat frequency) and major (coiling diameter) waves (Bressac et al. 1991b). In sperm monomorphic species, the minor and major waves are increased once they are transferred to the female (Bressac et al. 1991b). In sperm heteromorphic D. pseudoobscura, however, only the minor and major waves of long sperm increase within the female (Bressac et al. 1991b). Thus, if obscura group females use sperm motility to detect the number of sperm in storage, then female remating behaviour should be controlled by long, more motile sperm rather than short, less motile sperm. As stated previously, long sperm apparently play no role in female remating behaviour. In conclusion, no relationship was found between the actual number and proportion of long sperm in SSOs and female remating behaviour, and the only significant differences in sperm load between nonvirgin receptive and nonreceptive females was the proportion of each sperm type in the minor SSO, the ventral receptacle. Yet the most robust results in this study associated the probability of female remating with: (1) the absence of an egg in the uterus, and (2) a positive relationship between fecundity and productivity (and thus, the use of fertilizing sperm). The negative sperm load results, in conjunction with the positive progeny production results, suggest oviposition per se, rather than sperm, may be the primary influence determining female remating.
Acknowledgments This research was conducted in the laboratory of T. A. Markow and supported by a National Science Foundation grant DEB-9224263 and a Department of Zoology Graduate Student Research Fund. Special thanks goes to David Capco and Richard Satterlie for equipment use, Matt McGaughey and Matt DeAngelis for data collection, C. R. B. Boake for support during portions of the experimentation and T. L. Karr for support during portions of the writing. T. A. Markow, S. Arnold and J. B. Jones, Jr provided helpful discussion of the data and/or comments on the manuscript.
References Beatty, R. A. & Sidhu, N. S. 1970. Polymegaly of spermatozoan length and its genetic control in Drosophila species. Proceedings of the Royal Society of London, Series B, 71, 14–28. Bellis, M. A., Baker, R. R. & Gage, M. J. G. 1990. Variation in rat ejaculates consistent with the kamikaze-sperm hypothesis. Journal of Mammalogy, 71, 479–480. Birkhead, T. R. & Møller, A. P. 1992. Sperm Competition in Birds: Evolutionary Causes and Consequences. London: Academic Press. Bressac, C. & Hauschteck-Jungen, E. 1996. Drosophila subobscura females preferentially select long sperm for storage and use. Journal of Insect Physiology, 42, 323–328. Bressac, C., Joly, D., Devaux, J. & Lachaise, D. 1991a. Can we predict the mating pattern of Drosophila females from the sperm length distributions in males? Experientia, 47, 111–114. Bressac, C., Joly, D., Devaux, J., Serres, C., Feneux, D. & Lachaise, D. 1991b. Comparative kinetics of short and long sperm in sperm dimorphic Drosophila species. Cell Motility and the Cytoskeleton, 19, 269–274. Briskie, J. V. & Montgomerie, R. 1992. Sperm size and sperm competition in birds. Proceedings of the Royal Society of London, Series B, 247, 89–95. Brooks, D. R. & McLennan, D. H. 1991. Phylogeny, Ecology and Behavior. Chicago: University of Chicago Press. Cobbs, G. 1977. Multiple insemination and male sexual selection in natural populations of Drosophila pseudoobscura. American Naturalist, 111, 641–656. Cook, P. A. & Gage, M. J. G. 1995. Effects of risks of sperm competition on the numbers of eupyrene and apyrene sperm ejaculated by the moth Plodia interpunctella (Lepidoptera: Pyralidae). Behavioral Ecology and Sociobiology, 36, 261–268. Dickinson, J. L. 1986. Prolonged mating in the milkweed leaf beetle Labidomera clivicollis clivicollis (Coleoptera: Chrysomedlidae): a test of the ‘sperm-loading’ hypothesis. Behavioral Ecology and Sociobiology, 18, 331–338. Drummond, B. A. 1984. Multiple mating and sperm competition in the Lepidoptera. In: Sperm Competition and the Evolution of Animal Mating Systems (Ed. by R. L. Smith), pp. 291–370. London: Academic Press. Gage, A. R. & Barnard, C. J. 1996. Male crickets increase sperm number in relation to competition and female size. Behavioral Ecology and Sociobiology, 38, 349–353. Gage, M. J. G. 1991. Risk of sperm competition directly affects ejaculate size in the Mediterranean fruit fly. Animal Behaviour, 42, 1036–1037. Gage, M. J. G. 1994. Associations between body size, mating pattern, testis size and sperm lengths across butterflies. Proceedings of the Royal Society of London, Series B, 258, 247–254. Gage, M. J. G. & Cook, P. A. 1994. Sperm size or numbers? Effects of nutritional stress upon eupyrene and apyrene sperm production strategies in the moth Plodia interpunctella (Lepidoptera: Pyralidae). Functional Ecology, 8, 594–599. Gomendio, M. & Roldan, E. R. S. 1991. Sperm competition influences sperm size in mammals. Proceedings of the Royal Society of London, Series B, 243, 181–185. Gromko, M. H. & Markow, T. A. 1993. Courtship and remating in field populations of Drosophila. Animal Behaviour, 45, 253–262. Gromko, M. H., Newport, M. E. A. & Kortier, M. G. 1984. Sperm dependence of female receptivity to remating in Drosophila melanogaster. Evolution, 38, 1273–1282. Harvey, P. H. & Pagel, M. D. 1991. The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press. He, Y., Tanaka, T. & Miyata, T. 1995. Eupyrene and apyrene sperm and their numerical fluctuations inside the female reproductive tract of the armyworm, Pseudaletia separata. Journal of Insect Physiology, 41, 689–694.
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1507
Joly, D. & Lachaise, D. 1994. Polymorphism in the sperm heteromorphic species of the Drosophila obscura group. Journal of Insect Physiology, 40, 933–938. McLain, D. K. 1989. Prolonged copulation as a post-insemination guarding tactic in a natural population of the ragwort seed bug. Animal Behaviour, 38, 659–664. Manning, A. 1962. A sperm factor affecting the receptivity of Drosophila melanogaster females. Nature, 194, 252–253. Miller, A. 1950. The internal anatomy and histology of the imago of Drosophila melanogaster. In: Biology of Drosophila (Ed. by M. Demerec), pp. 420–534. New York: J. Wiley. Parker, G. A. 1970. Sperm competition and its evolutionary consequences in the insects. Biological Review, 45, 525–567. Parker, G. A. & Simmons, L. W. 1994. Evolution of phenotypic optima and copula duration in dungflies. Nature, 370, 53–56. Pasini, M. E., Redi, C. V., Caviglia, O. & Perotti, M. E. 1996. Ultrastructual and cytochemical analysis of sperm dimorphism in Drosophila subobscura. Tissue and Cell, 28, 165–175. Patterson, J. T. 1954. II. The fate of the sperm in the reproductive tract of the Drosophila female in homogamic matings. University of Texas Publication, Genetics, 5422, 19–37. Pitnick, S. 1996. Investment in testes and the cost of making long sperm in Drosophila. American Naturalist, 148, 57–80. Pitnick, S. & Markow, T. A. 1994a. Male gametic strategies: sperm size, testes size, and the allocation of ejaculate among successive mates by the sperm-limited fly Drosophila pachea and its relatives. American Naturalist, 143, 785–819. Pitnick, S. & Markow, T. A. 1994b. Large male advantages associated with costs of sperm production in Drosophila hydei, a species with giant sperm. Proceedings of the National Academy of Sciences, U.S.A., 91, 9277–9281. Pitnick, S., Markow, T. A. & Spicer, G. S. 1995. Delayed male maturity is a cost of producing large sperm in Drosophila. Proceedings of the National Academy of Sciences, U.S.A., 92, 10 614–10 618. Pruzan-Hotchkiss, A., Dejianne, D. & Faro, S. H. 1981. Sperm utilization in once- and twice-mated Drosophila pseudoobscura females. American Naturalist, 118, 37–45. Pyle, D. W. & Gromko, M. H. 1978. Repeated mating by female Drosophila melanogaster: the adaptive importance. Experientia, 34, 449–450. Sanger, W. G. & Miller, D. D. 1973. Spermatozoan length in species of the Drosophila affinis subgroup. American Midland Naturalist, 90, 485–489. Schwartz, J. M. & Boake, C. R. B. 1992. Sexual dimorphism in remating in Hawaiian Drosophila species. Animal Behaviour, 44, 231–238.
Silberglied, R. E., Shepherd, J. G. & Dickinson, J. L. 1984. Eunuchs: the role of apyrene sperm in Lepidoptera? American Naturalist, 123, 255–265. Simmons, L. W., Craig, M., Llorens, T., Schinzig, M. & Hosken, D. 1993. Bushcricket spermatophores vary in accord with sperm competition and parental investment theory. Proceedings of the Royal Society of London, Series B, 251, 183–186. Siva-Jothy, M. T. 1987. Variation in copulation duration and the resultant degree of sperm removal in Orthetrum cancellatum (L.) (Libellulidae: Odonata). Behavioral Ecology and Sociobiology, 20, 147–151. Sivinski, J. 1980. Sexual selection and insect sperm. Florida Entomologist, 63, 99–111. Smith, R. L. 1984. Sperm Competition and the Evolution of Animal Mating Systems. London: Academic Press. Snook, R. R. 1995. The evolution of sperm polymorphism in the Drosophila obscura group. Ph.D. thesis, Arizona State University. Snook, R. R. 1997. Is the production of multiple sperm types adaptive? Evolution, 51, 797–808. Snook, R. R. & Karr, T. L. 1998. Only long sperm are fertilizationcompetent in six sperm-heteromorphic Drosophila species. Current Biology, 8, 291–294. Snook, R. R. & Markow, T. A. 1996. Possible role of nonfertilizing sperm as a nutrient source for female Drosophila pseudoobscura (Diptera: Drosophilidae). Pan-Pacific Entomologist, 72, 121–129. Snook, R. R., Markow, T. A. & Karr, T. L. 1994. Functional nonequivalence of sperm in Drosophila pseudoobscura. Proceedings of the National Academy of Sciences, U.S.A., 91, 11 225–11 229. Thiobold, E. 1979. Stimulation of reproductive activity of females of Acrolepiopsis assectella (Lepidoptera: Hyponomeutoidae) by the presence of eupyrene spermatozoa in the spermatheca. Entomologia Experimentalis et Applicata, 26, 279–290. Tompkins, L. & Hall, J. C. 1983. Identification of brain sites controlling female receptivity in mosaics of Drosophila melanogaster. Genetics, 103, 179–195. Trevitt, S., Fowler, K. & Partridge, L. 1988. An effect of eggdeposition on the subsequent fertility and remating frequency of female Drosophila melanogaster. Journal of Insect Physiology, 34, 821–828. Wilkinson, L. 1990. SYSTAT: the System for Statistics. Evanston, Illinois: SYSTAT. Wedell, N. 1992. Protandry and mate assessment in the wartbiter Decticus verrucivorus (Orthoptera: Tettigoniidae). Behavioral Ecology and Sociobiology, 31, 301–308. Yasui, Y. 1994. Adaptive control of copulation duration by males under sperm competition in the mite, Macrocheles muscaedomesticae. Experimental and Applied Acarology, 18, 543–554.
The risk of sperm competition and the evolution of sperm heteromorphism RHONDA R. SNOOK
Department of Zoology, Arizona State University, Tempe (Received 3 February 1998; initial acceptance 27 February 1998; final acceptance 30 March 1998; MS. number: A8025R)
ABSTRACT Members of the Drosophila obscura species group exhibit sperm heteromorphism in which males simultaneously produce two different morphologies of sperm, short and long. Short sperm represent at least 50% of the ejaculate in several species but do not function in fertilization and thus, the evolutionary significance of this phenomenon is unknown. Selective pressures associated with the risk of sperm competition have been suggested to explain the maintenance of sperm heteromorphism. Here I test: (1) whether males alter their behaviour or ejaculate characteristics in response to the risk of sperm competition, and (2) whether short, nonfertilizing sperm serve as ‘cheap filler’ in the female reproductive tract, thereby delaying female receptivity to subsequent matings. Males did not alter tactics based on the risk of sperm competition; copulation durations of males mated to nonvirgin females were unexpectedly shorter than when both sexes were virgin and, contrary to sperm competition predictions, males did not alter the ratio of fertilizing:nonfertilizing sperm relative to female mating status or female age. Short sperm had a small, if any, role in influencing female remating behaviour. The number and proportion of short sperm present in sperm storage organs did not differ between nonvirgin receptive and nonvirgin nonreceptive females. The proportion of short sperm present in the ventral receptacle, however, was lower in receptive than nonreceptive females. Remating behaviour was strongly linked to oviposition patterns. Nonvirgin receptive females oviposited more eggs than nonreceptive females, and female remating interval was positively related to both the number of eggs and progeny produced. Although, oviposition parameters are related to fertilization and thus, the use of long sperm, nonvirgin receptive and nonreceptive females did not predictably differ in the number or proportion of long sperm in storage. These results suggest oviposition per se is probably more important than sperm in determining female receptivity and that sperm heteromorphism may play a marginal, if any, role in affecting female remating.
1994; Snook 1995; Bressac & Hauschteck-Jungen 1996). Both sperm types are nucleated, have the same DNA content (Pasini et al. 1996) and are motile, suggesting both types of sperm can fertilize eggs. Direct measurements of sperm in eggs, however, found that only long sperm fertilize eggs in several obscura group species (D. pseudoobscura, Snook et al. 1994; D. persimilis, D. affinis, D. miranda, D. subobscura and D. athabasca, Snook & Karr 1998; and probably D. obscura, Snook 1997). The evolutionary significance of sperm heteromorphism is unknown. Interpreting the functional significance of this phenomenon in the obscura group is confounded by the fact that all obscura species so far examined produce two discrete types of sperm (Beatty & Sidhu 1970; Sanger & Miller 1973; Joly & Lachaise 1994; Snook 1997), whereas no other Drosophila species exhibits
Males of many species produce multiple sperm types within each ejaculate, a phenomenon known as sperm heteromorphism (Sivinski 1980; Silberglied et al. 1984). For example, lepidopterans produce both small, apyrene (anucleated) sperm, which do not function in fertilization, and larger eupyrene (enucleate) fertilizing sperm (Silberglied et al. 1984). Males of the Drosophila obscura group produce two types of mature sperm, ‘short’ sperm which have short heads and tails and ‘long’ sperm which have long heads and tails (Beatty & Sidhu 1970; Sanger & Miller 1973; Joly & Lachaise 1994; Snook 1997). Short sperm represent at least 50% of the ejaculate (Snook et al. Correspondence and present address: R. R. Snook, Department of Biological Sciences, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154-4004, U.S.A. (email: [email protected]). 0003–3472/98/121497+11 $30.00/0
1998 The Association for the Study of Animal Behaviour
1497
1998 The Association for the Study of Animal Behaviour
1498 ANIMAL BEHAVIOUR, 56, 6
this phenomenon. The production of multiple sperm types in any particular species, therefore, may reflect common ancestry rather than adaptive function (although similarity by descent may be selectively maintained; Brooks & McLennan 1991; Harvey & Pagel 1991). Recent phylogenetic autoregression analyses, however, revealed that 22% of the interspecific variation in the length of long sperm is related to phylogeny, whereas the length of short sperm is not significantly correlated with phylogeny, indicating different selection pressures have acted on the two sperm lengths (Snook 1997). Long sperm have evolved in response to fertilization demands, whereas short sperm have been decoupled from these requirements and may be an adaptation to currently unidentified selection pressures (Snook 1997; Snook & Karr 1998). Several selection pressures have been suggested to explain the maintenance of nonfertilizing gametes, namely that nonfertilizing sperm serve as nutrient contributions (Gage & Cook 1994; Cook & Gage 1995; Snook & Markow 1996) and nonfertilizing sperm function in sperm competition (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995; Snook & Karr 1998), where more than one male’s ejaculate competes within the female reproductive tract for fertilization opportunities (Parker 1970). In both lepidopterans and Drosophila, nonfertilizing sperm do not function as nutrient contributions (Cook & Gage 1995; Snook & Markow 1996). Nonfertilizing sperm in lepidopterans may function in sperm competition (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995), but sperm competition hypotheses have not been tested in the obscura group. Both male behaviour (i.e. copulation duration) and ejaculate characteristics may be influenced by the perceived risk of sperm competition (Smith 1984; Birkhead & Møller 1992). As copulation duration increases, paternity tends to increase. Males of certain species prolong copulation with previously mated females, allowing males to increase paternity by: (1) delivering more sperm (Dickinson 1986), (2) mate guarding (Parker 1970; McLain 1989), or (3) volumetrically removing more of a rival male’s sperm from the female sperm storage organs (Siva-Jothy 1987). Ejaculate characteristics, such as the number of sperm, or in sperm heteromorphic systems, the proportion of sperm types, can also change with the risk of sperm competition. Males may alter the number and/or type of sperm transferred to females in response to: (1) female mating status (Smith 1984; Birkhead & Møller 1992; Wedell 1992), (2) presence of rival males (Gage 1991; Simmons et al. 1993; Gage & Barnard 1996), or (3) female age (Cook & Gage 1995; Gage & Barnard 1996). For example, in sperm heteromorphic lepidopterans, males transfer more fertilizing sperm to previously mated females and transfer fewer nonfertilizing sperm to older females (Cook & Gage 1995). These results indicate that males increase their reproductive success by transferring larger numbers of fertilizing sperm to compete with a prior male’s sperm and suggest that, because younger females are likely to have higher future fecundity, males transfer more nonfertilizing sperm to delay female
remating behaviour and prolong the use of their sperm (Cook & Gage 1995). The effect of sperm competition on ejaculate and behavioural traits in sperm heteromorphic obscura group species is not known. Nonfertile sperm morphs may play a specific role in sperm competition by either functioning to delay female remating by serving as ‘cheap filler’ within female sperm storage organs (SSOs) or displacing/incapacitating rival sperm already in storage (Silberglied et al. 1984; Gage & Cook 1994; Cook & Gage 1995). Several aspects of Drosophila biology suggest that short, nonfertilizing sperm could function in sperm competition by serving as cheap filler to delay female remating. First, in all obscura group species examined, females multiply mate so that sperm from more than one male competes to fertilize eggs (Cobbs 1977; Pruzan-Hotchkiss et al. 1981). Second, in many Drosophila species, as with lepidopterans, long-term inhibition of female remating is linked to sperm load (Manning 1962; Drummond 1984; Gromko et al. 1984; Schwartz & Boake 1992; Gromko & Markow 1993; Pitnick & Markow 1994a; He et al. 1995), either through the positive correlations between female remating behaviour and progeny production, an indicator of sperm use (Trevitt et al. 1988; Gromko & Markow 1993), or between the number of sperm received per copulation and female remating interval (Pitnick & Markow 1994a). Third, short sperm are probably relatively cheaper to produce than long sperm as Drosophila males that produce longer sperm mature later than males that produce shorter sperm (Pitnick et al. 1995). Here I test the hypothesis that sperm heteromorphism is an adaptation to sperm competition by assessing whether: (1) males alter their copulation duration or ejaculate characteristics in response to the risk of sperm competition, and (2) short sperm serve as cheap filler delaying female receptivity (Silberglied et al. 1984). METHODS
Fly Maintenance T. Markow collected D. pseudoobscura in February 1991 from fallen citrus in Tempe, Arizona, D. affinis (14012–0141.2) was obtained from the National Drosophila Species Resource Center, Bowling Green, Ohio, and D. persimilis was kindly provided by A. T. Beckenbach (Simon Fraser University, Burnaby, British Columbia). All cultures were maintained on standard cornmeal–agar–molasses food with yeast and kept at room temperature, 22–25C, and an approximate 12:12 h light:dark cycle. I collected virgin flies, separated males and females upon eclosion by aspiration, and stored flies 10 per 10-dram yeasted food vials. All flies used were 5 days old, unless otherwise noted, and reproductively mature (Snook 1995).
Sperm Counts I determined the ratio or number of sperm types by dissecting sperm from either male seminal vesicles or
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1499
female reproductive tracts in phosphate-buffered saline (PBS) on a microscope slide, fixing the sperm, and then staining the sperm with Hoescht’s 33258, a DNA-specific dye, as described previously (Snook et al. 1994). I then counted sperm on an epifluorescent microscope. Males produce and transfer to females too many sperm to enumerate completely so I counted 500 sperm per sample to get proportion data for seminal vesicles and uterus sperm counts. Sperm counts in SSOs were based on total enumeration because fewer sperm are found in these organs (Snook et al. 1994).
Copulation Duration, Female Mating Status, Female Age and Sperm Transfer Patterns I performed interrupted copulations on D. pseudoobscura to determine whether the two sperm types were transferred at different times during copulation. This was assessed as a preliminary examination of the role sperm length may play in the timing of the entrance into SSOs and the competitive ability of different sperm lengths, two parameters previously invoked as determinants of the outcome of sperm competition (Gomendio & Roldan 1991; Briskie & Montgomerie 1992; Pitnick & Markow 1994a). I paired two virgin males and one virgin female and interrupted copulations at 30-s intervals up to 3 min by either gently aspirating the mating pair or tapping the vial. I dissected females immediately after copulation was interrupted and determined the proportion of sperm types in the uterus. I examined the influence of female mating status on copulation durations in three species, D. pseudoobscura, D. persimilis and D. affinis. Males mated to nonvirgin females are predicted to have longer copulation durations than when mated to virgin females. I paired two virgin males and one virgin female for mating and subsequently exposed mated females to virgin males for 2 h in the morning and afternoon on each of 5 consecutive days (periodic interaction design; Pyle & Gromko 1978). I recorded durations of all copulations. Additionally, I examined the influence of male mating status on copulation duration in two species, D. pseudoobscura and D. persimilis. For this experiment, I paired two virgin females and one virgin male. I removed the females when copulation had ended and replaced them with two new virgin females until the next copulation occurred. I recorded the durations of all copulations. To determine whether female mating status influenced the proportion of sperm types males transferred, I paired virgin D. pseudoobscura males to nonvirgin females that had mated 48 h earlier. Immediately following the females’ second copulation, I dissected females and determined the proportion of sperm the second male transferred (sperm storage does not begin until ca. 2 h after copulation). I then compared these data with the proportion of sperm transferred by virgin males to virgin females (data from Snook et al. 1994). To determine whether female age influenced the proportion of sperm types transferred to females, I paired virgin males and virgin females that were either 5 or 15 days old. Immediately following copulation, I dissected
the females and determined the proportion of sperm transferred.
The Cost of Sperm Production For short sperm to function as cheap filler, they should be less costly to produce than long sperm. One measurement of cost is the rate at which males produce each sperm type. I paired one set of D. pseudoobscura males to virgin females once every day for 15 days (Daily mating treatment), and paired another set of males to virgin females once every other day (Alternate mating treatment). I discarded data for individual males if the males failed to remate anytime during the experiment. Finally, I maintained another set of virgin males and did not pair them with females for 15 days (Virgin 15-day treatment). On the 16th day of the experiment, I dissected all males and processed the seminal vesicles for sperm counting. I maintained all females that mated with males during Daily or Alternate mating treatments in individual vials, and transferred the females to new food vials every other day for 5 days. I retained all vials for subsequent progeny counts and compared the number of progeny produced between females mated to males of the two mating groups.
Sperm Load, Female Receptivity and the Cheap Filler Hypothesis To determine whether short sperm increase female remating latency, I tested the prediction that nonvirgin receptive females should have fewer short sperm (less filler) than nonvirgin nonreceptive females. I paired two virgin D. pseudoobscura males with a virgin female in a yeasted food vial. Upon mating, I removed the nonmating male from the vial by aspiration and, following the completion of copulation, removed the mating male. Thirty-six hours later, I paired two virgin males with a previously mated female in a yeasted food vial for a maximum of 2 h. Thirty-six hours was chosen because the average D. pseudoobscura female remating interval is 2 days (48 h; see below). A female was considered receptive to remating if copulation occurred (the male mounted and inserted his genitalia into the female). When this copulation occurred, I immediately aspirated the pair apart prior to sperm transfer, removed the male from the vial, and immediately dissected the female for sperm counts. If females did not indicate a willingness to remate by the end of the 2-h receptivity test, they were conidered nonreceptive. In these cases, I removed males from the vials, and immediately dissected the nonreceptive females for sperm counts. I also counted the number of eggs oviposited in vials that held females during the time between the first mating and the receptivity test to determine sperm use prior to remating. I discarded females that died on or before day 5 of the experiment. I used one-tailed t tests to determine whether nonvirgin nonreceptive females and nonvirgin receptive females differed in the sperm types stored because the prediction specifically states that the mean number or proportion of sperm types stored by nonvirgin
1500 ANIMAL BEHAVIOUR, 56, 6
nonreceptive females should be greater than receptive females (ìnonreceptive >ìreceptive).
1.0 1/8
Sperm Use and Female Remating Patterns
RESULTS
Copulation Durations, Female Mating Status, Female Age and Sperm Transfer Patterns Virgin males did not begin to transfer sperm to virgin females until approximately 1.5 min into copulation (Fig. 1). Both sperm types were transferred to females at the start of sperm transfer (Fig. 1). Furthermore, the proportion of short sperm transferred by males did not differ, either during any 30-s interruption of copulation or from complete copulations (ANOVA: F4,30 =0.264, P=0.9; Fig. 1). Female and male mating status influenced copulation duration. In all three species examined, the copulation durations of males mated to previously mated females were significantly less than that of males mated to virgin females (Table 1). Similarly, copulation durations of females mated to previously mated males were significantly less than that of females mated to virgin males in D. pseudoobscura and D. persimilis (D. affinis was not tested; Table 1). Female mating status did not influence the ratio of sperm types males transferred to females. The proportion of short sperm transferred by virgin males to virgin females (0.4790.023, N=11) did not differ from the proportion transferred to females that had mated 48 h earlier (0.5080.029, N=13: t22 =0.761, P=0.45). Female age did not influence the ratio of sperm types transferred to females. The proportion of short sperm
0.8 Proportion of short sperm
To estimate the relationship between female remating and sperm use, I examined the correlations between egg and progeny production and remating interval in D. pseudoobscura, D. persimilis and D. affinis. I allowed females the opportunity to mate for 2 h in the morning and afternoon for 5 consecutive days. I transferred females to new food vials containing dried yeast paste once a day and/or after remating for the first 15 days, and then every other day until death. I retained all vials for subsequent adult progeny counts. I discarded data for individual females if the females did not remate or failed to produce progeny. I transferred singly mated control females to new food vials and treated them as described for multiply mated females. I performed two to six replicates of each treatment, either multiply mated or singly mated females, in each species. I tested replicates for homogeneity using independent t tests or ANOVA. If the replicates were homogeneous then I pooled data. I subsequently performed a linear regression analysis to describe the relationship between female remating and the production of eggs and progeny. I also compared the total number of eggs oviposited (fecundity) and progeny produced (productivity) by singly or multiply mated females. I performed all statistical analyses using SYSTAT (Wilkinson 1990).
0.6
4/9
12/14 2/9
0.4
11/11
6/11
0.2
0/6
0
0.5
1.0
1.5 2.0 2.5 3.0 Copulation duration (min)
6.0*
Figure 1. The proportion (±SE) of short sperm found in the uterus of females following the interruption of copulation at 30-s intervals up to 3 min and a full copulation. *Data from Snook et al. (1994). Numbers in bars indicate the number of males that transferred sperm (numerator) in the number of pairs tested (denominator).
transferred by virgin males to 5-day-old virgin females (0.4730.013, N=20) did not differ from the proportion transferred to 15-day-old virgin females (0.4770.017, N=19: t37 =0.838, P=0.84).
The Cost of Sperm Production Male mating behaviour influenced sperm production patterns in that males given either daily or alternate (every other day) opportunities to mate produced a significantly higher proportion of short sperm than virgin males of any age examined (F3,47 =8.380, P=0.0001; Fig. 2). The proportion of short sperm produced by males that mated daily did not differ from that produced by males that mated every other day (Fig. 2). The number of progeny produced by females that mated to either daily mated males or alternately mated males did not differ (t20 =1.794, P=0.09; Fig. 2).
Sperm Load, Female Receptivity and the Cheap Filler Hypothesis Nonvirgin receptive D. pseudoobscura females oviposited more eggs, suggesting they had used more sperm, than nonvirgin nonreceptive females (t103 = 2.336, P=0.01; Fig. 3). Moreover, 50 of 65 nonreceptive females had an egg in the uterus upon dissection, whereas 63 of 66 receptive females did not (G=81.18, P<0.001). Despite these apparent differences in sperm use, nonvirgin receptive and nonreceptive females had the same total number of sperm in storage (irrespective of sperm type and sperm
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1501
Table 1. The copulation duration of virgin females mated to virgin males, and those same females when they remated, and the copulation duration of virgin males mated to virgin females, and those same males when they remated, and the sample size (N), paired t, and P for each comparison of three obscura group species Copulation duration (min)
Copulation duration (min)
Species
Virgin /
Remating /
N
Paired t
P
Virgin ?
Remating ?
N
Paired t
P
D. pseudoobscura D. persimilis D. affinis
6.2±0.2 7.9±0.3 1.4±0.1
3.7±0.2 5.2±0.3 1.1±0.1
52 45 24
8.597 5.374 2.598
<0.001 <0.001 0.02
6.4±0.2 5.4±0.3
4.4±0.4 4.6±0.4
43 22
7.141 2.047
<0.001 0.05
storage location; t46 =1.191, P>0.1; Fig. 3), suggesting total sperm load was not related to female remating behaviour. No significant differences in the number of either short or long sperm were observed between nonvirgin receptive and nonreceptive females in any comparison (Fig. 4). The proportion of either short or long sperm in the spermathecae and all SSOs also did not differ between females (Fig. 4). The proportion of long sperm in the ventral receptacle of nonreceptive females, however, was significantly less than that of receptive females, contrary to predictions that sperm load influences female remating behaviour. The proportion of short sperm in the ventral receptacle of nonreceptive females was significantly greater than that of receptive females, supporting the
0.75 B B
hypothesis that short sperm serve as cheap filler (Fig. 4).
Sperm Use and Female Remating Patterns Female D. pseudoobscura usually mated three times during the 5-day mating portion of the experiment (median number of mates=3, total number remating=50/60), and the median remating interval from the initial mating to the first remating was 2.0 days. Females oviposited 69.4 (4.5) eggs and produced 52.7 (3.7) progeny between the initial mating and the first remating. There was a significant positive correlation between remating interval and the number of either eggs or progeny produced (Fig. 5a). Of the 50 females that initially remated, 39 remated again. In this case, the median remating interval was 1.5 days between the first remating and the second remating. Between the first and second remating, females oviposited 47.0 (4.6) eggs and produced 33.9 (3.7) progeny. There was also a significant positive correlation
11
0.50
A
1300
13 16
1200
Nonreceptive Receptive
1000
0.25
0.00
24 24
1100
Number
Proportion of short sperm
11
A
Virgin 5 days
Virgin 15 days
Daily
Alternate
900 800 700 200 P < 0.05
Treatment Figure 2. The proportion (±SE) of short sperm found in the seminal vesicles of 5- and 15-day-old virgin males, and in males given the opportunity to either mate every day for 15 days (Daily), or every other day for 15 days (Alternate). Different letters denote significant differences between groups. Females in the Daily mating treatment produced 601.5±31 progeny, and females in the Alternate mating treatment produced 683±26 progeny. Numbers in bars indicate the sample size.
100
0
54
51
Eggs
Total sperm
Figure 3. The mean (±SE) number of eggs and total number of sperm (short+long) found in all SSOs of nonreceptive and receptive females. Numbers above bars are sample sizes. See text for statistics.
1502 ANIMAL BEHAVIOUR, 56, 6
200
Nonreceptive Receptive
t 47 = –1.351
(a) 175
0.25 t 47 = 1.677, P = 0.05
150 0.20
125 100
0.15
75
0.10 t 47 = 0.887
50
0.05
25 24
25
0
1000 900
t 47 = 0.979
(b)
0.20 t 47 = 0.816
700
Proportion of short sperm
Number of sperm
800
600 500 400 300
t 47 = 1.289 24
200
25
100 0
0.15
0.10
0.05
0
1200 1100
(c)
t 46 = 0.966
1000
0.20
900
t 46 = 0.981
800 0.15
700 600 500
0.10
400 300 200
t 46 = 1.485 24
0.05 24
100 0
Long
Short
0
Sperm type Figure 4. The number of short and long sperm and the proportion of short sperm found in the (a) ventral receptacle, (b) spermathecae and (c) all SSOs, in nonvirgin nonreceptive and receptive females. Numbers above bars are sample sizes for both sperm types. All comparisons, except the proportion of short sperm in the ventral receptacle, were NS at P>0.05.
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1503
140 120
140 (a) N = 50
120
100
100
80
80
60
60
40
40
20
20
0
0
(b) N = 39
120 (c) N = 50 Number of eggs and progeny
100
80
60
40
20
0 100
100 (d) N = 26
(e) N = 13
80
80
60
60
40
40
20
20
0
1
2
3
4 0 Remating interval (days)
1
2
3
4
Figure 5. The relationship between the number of eggs oviposited and progeny produced either between the initial mating and the first remating (a, c, d), or between the first and second remating (b, e), and the remating interval for D. pseudoobscura (a, b), D. persimilis (c), and D. affinis (d, e). Remating interval was calculated as the number of days between one mating and the next. (a) Eggs: R2 =0.41, P<0.001; progeny: R2 =0.39, P<0.001, (b) eggs: R2 =0.26, P=0.001; progeny: R2 =0.18, P<0.01, (c) eggs: R2 =0.18, P<0.01; progeny: R2 =0.16, P<0.01, (d) eggs: R2 =0.32, P<0.01; progeny: R2 =0.312, P<0.05, (e) eggs: R2 =0.53, P<0.01; progeny: R2 =0.17, P>0.05. x: eggs; _: progeny.
1504 ANIMAL BEHAVIOUR, 56, 6
in fecundity and productivity between the first and second remating (Fig. 5b). There was no significant difference in the average time to remating between the initial–first remating and the first–second rematings for D. pseudoobscura females (paired t38 =1.188, P=0.2). Females produced fewer progeny between the first and second remating interval compared with the number produced between the initial and first remating interval (paired t38 =2.118, P=0.04), but fecundity did not differ between remating bouts (paired t38 =1.904, P=0.06). Female D. persimilis tended to remate only once during the remating test (median number of mates=2, total number of females remating=50/60, median remating interval=2.375 days). Irrespective of remating intervals, D. persimilis females oviposited 55.8 (3.2) eggs and produced 30.7 (3.0) progeny between the initial mating and the first remating. This approximately 50% reduction in the number of eggs oviposited compared with progeny produced may be a function of unsuccessful fertilizations (recent analyses have found that half the eggs oviposited are not fertilized; Snook & Karr 1998). There was a significant positive correlation between remating interval and both the number of eggs and progeny produced (Fig. 5c). Of the 50 females that initially remated, only five remated again, therefore I did not analyse the association between the first and second remating interval and fecundity and productivity. Half the D. affinis females remated once, and the other half remated twice (median number of mates=2, total number of females remating=28/30, median remating interval=1.5 days). Females oviposited 47.5 (3.2) eggs and produced 22.8 (2.8) progeny between the initial mating and the first remating. As with the other two species, there was a significant positive correlation between the initial remating interval and the number of eggs and progeny produced (Fig. 5d). Of the 13 females that remated twice, the median remating interval was 2.0 days after the initial remating. These females oviposited 47.8 (6.1) eggs and produced 17.3 (3.2) progeny between the first and second remating. Again, there was a significant positive correlation between remating interval and egg and progeny production (Fig. 5e). Female D. affinis took significantly longer to remate the second time than they did between their initial mating and first remating (paired t12 =2.843, P=0.01). Fecundity and productivity, however, did not differ between the initial mating and the first remating and the first and second remating intervals (eggs: paired t12 =0.421, P=0.7; progeny: paired t12 =0.76, P=0.5).
DISCUSSION
Alteration of Behavioural and Ejaculate Characteristics Ejaculate characteristics and male behaviour in many species, including those that are sperm heteromorphic, are influenced by the perceived risk of sperm competition (Smith 1984; Bellis et al. 1990; Gage 1991; Birkhead &
Møller 1992; Parker & Simmons 1994; Cook & Gage 1995). However, obscura group males do not alter ejaculate (proportion of short and long sperm) or behavioural (copulation duration) characteristics in response to a perceived risk of sperm competition. Drosophila pseudoobscura males transfer both sperm types simultaneously during copulation to virgin females and both sperm types enter SSOs at approximately the same time (Snook et al. 1994; Snook 1995). These results indicate that sperm length per se does not influence the timing of sperm entrance into SSOs and suggests that sperm length may not influence sperm competitive ability as has been proposed for sperm monomorphic systems (Gomendio & Roldan 1991; Briskie & Montgomerie 1992). Additionally, these results suggest that short sperm are not ‘front runners’, which serve to actively displace a previous male’s sperm (Silberglied et al. 1984), because both sperm types are transferred and stored simultaneously. Males also do not alter the proportion of short and long sperm based on female mating status. In lepidopterans, males also do not increase the number of nonfertilizing sperm based on female mating status, but do increase the absolute number of fertilizing sperm (Cook & Gage 1995). Although I did not measure the absolute number of short and long sperm transferred to females of different mating status, I found no proportional changes. Male D. pseudoobscura transfer approximately 50% of each sperm type to females; therefore, any potential differences in the absolute number of sperm transferred would have been equal for both types of sperm. In the Indian meal moth, Plodia interpunctella, males transfer more nonfertilizing sperm to young females compared with older females (Cook & Gage 1995). This difference may ultimately be a result of younger females having a longer potential life span, thus a higher reproductive value and increased risk of encountering other males, which males attempt to counter by transferring more ‘cheap filler’ to younger females (Cook & Gage 1995). However, in D. pseudoobscura, female age did not influence the proportion of each sperm type transferred by females. Five-day-old virgin females received the same proportion of short and long sperm as did 15-day-old virgin females. Predictions about the causal factors influencing variation in copulation duration have been described for some species. These factors are complex and can depend on the form of sperm precedence (Yasui 1994), female mating status and oviposition patterns (Siva-Jothy 1987), size of males (Parker & Simmons 1994), and age of males (Siva-Jothy 1987). In general, however, longer copulations lead to a higher reproductive success for males. In the three obscura group species examined, males that mated with nonvirgin females and thus, experienced sperm competition, copulated for an unexpected shorter duration than males that mated with virgin females. This supports a previous study examining copulation duration in D. affinis (Bressac et al. 1991a). In D. pseudoobscura and D. persimilis, male mating status also influenced copulation duration. Mated males copulated for a shorter duration with virgin females than did virgin males. Thus, under the scenarios tested, I found no evidence that
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1505
males respond predictably to sperm competition risks through ejaculate or behavioural changes.
The Cheap Filler Hypothesis Studies on Drosophila that produce giant sperm have found that long sperm are costly to produce (Pitnick & Markow 1994b; Pitnick et al. 1995; Pitnick 1996), thus intuitively, within the obscura group, short sperm are likely to be less costly to produce than long sperm. Here I found that frequent mating by D. pseudoobscura males results in males producing a lower proportion of long sperm compared with virgin males of any age. These results suggest that it takes longer to produce long sperm, and thus, reflects a two-fold cost to producing long sperm: increased time to produce, and when frequently mating, a decrease in the number of long sperm available for fertilization. Short sperm are cheaper to produce than long sperm, but do they function as filler? To function as filler, short sperm must influence female remating behaviour. Previous studies linked female remating behaviour to sperm load by examining the relationship between progeny production and female remating (e.g., Gromko et al. 1984; Schwartz & Boake 1992; Gromko & Markow 1993; He et al. 1995). I found a significant positive correlation between remating intervals and both fecundity and productivity in all three obscura species examined. Female D. melanogaster that oviposit more eggs or produce more progeny tend to remate before females that oviposit or produce fewer eggs or progeny, and this evidence has been used to link sperm load and female remating behaviour (Trevitt et al. 1988). Moreover, field-caught D. melanogaster and D. simulans females that remate produce fewer progeny than females that do not remate, suggesting that females wait to remate until ‘a substantial number of sperm’ from a previous mating have been used (Gromko & Markow 1993). Here I found similar results: receptive females oviposited more eggs, and thus used more sperm, than nonreceptive females. If remating interval is determined by sperm load, however, then the number of eggs oviposited and/or progeny produced between the initial mating and the first remating and the first and second remating intervals should be the same, assuming males transfer and/or females store similar numbers of sperm after each mating. (I did not directly test this assumption, however, males do not alter the proportion of sperm types transferred to females.) Contrary to this prediction, the productivity between the initial and first remating was significantly greater than the productivity between the first and second rematings for D. pseudoobscura. Additionally, even if a second male transfers more sperm to a mated female to better compete with a prior male’s sperm, productivity between the first and second remating should be higher, rather than lower as was found. In D. affinis, females oviposited the same number of eggs and produced the same number of progeny between remating episodes despite the second remating interval being significantly longer than the first, also indicating sperm use had little affect on remating interval. Other studies found discrepancies between the
number of females remating based on oviposition patterns and the association with sperm load. For example, Pruzan-Hotchkiss et al. (1981) found that there was no difference in the proportion of D. pseudoobscura females remating at 3 and 6 days after their initial mating despite a three-fold difference in stored sperm use. If sperm load determined remating, then a higher proportion of females would be expected to remate at 6 days compared with 3 days. These combined data suggest sperm load does not effect female remating in D. pseudoobscura. The confounding element in these results, however, is that indirect fecundity and productivity data only address the influence of long sperm on remating interval because only long sperm are used for fertilization (Snook et al. 1994; Snook & Karr 1998). In the present study, there was no difference between the number and proportion of long sperm found in SSOs of nonvirgin receptive and nonreceptive females, which supports the notion that long sperm do not mediate female remating interval. In lepidopterans, Cook & Gage (1995) found males transfer more nonfertilizing sperm to younger females that are more likely to remate compared with older females. Additionally, the number of nonfertilizing sperm in the armyworm, Pseudaletia separata, decreases within 3 days after copulation, with most disappearing within the first 24 h, and this disappearance correlates with female remating behaviour (He et al. 1995). Similarly, in all three obscura group species examined here, the number of short sperm in SSOs was severely reduced 48 h after copulation (Snook et al. 1994; Snook 1995). The significance of this reduction is unknown; however, the reduction does not reflect their use in fertilization (Snook et al. 1994; Snook & Karr 1998) or the incorporation of nutrient donations (Snook & Markow 1996), nor does it appear to influence female remating behaviour. In the present study, nonvirgin receptive and nonreceptive females did not differ in the number of short sperm stored in any SSO. The proportion of short sperm stored in the ventral receptacle of nonreceptive females was significantly greater than in receptive females, but the proportion of short sperm in the spermathecae and the SSOs combined were not significant. Most D. pseudoobscura sperm are found in the spermathecae (Snook et al. 1994) and sperm from all SSOs are used equally (Patterson 1954). Given this, if short sperm influence female remating behaviour, then nonvirgin nonreceptive females would be expected to have more filler in the spermathecae or both SSO types, contrary to the proportional difference found only in the ventral receptacle.
Sperm Motility, Oviposition and Female Remating Behaviour In lepidopterans, the presence of sperm in the spermatheca delays female remating behaviour (Drummond 1984), and this behavioural modification is thought to be mediated by sperm motility (Thiobold 1979). Nonfertilizing sperm are initially more motile than eupyrene sperm so they enter SSOs first, subsequently generating motility
1506 ANIMAL BEHAVIOUR, 56, 6
and delaying female remating (Silberglied et al. 1984; Cook & Gage 1995; He et al. 1995). Additionally, Gage (1994) found no relationship between the risk of sperm competition and nonfertilizing sperm length, but this length did correlate with male body size. Gage (1994) suggested that if body size correlates with female reproductive tract morphology, then nonfertilizing sperm may function as cheap filler in SSOs through motility differences between sperm types. Female D. melanogaster receptivity is proximately controlled by the central nervous system (Tompkins & Hall 1983). A mechanical stimulation of the central nervous system through sperm motility has been suggested to influence female remating (Gromko et al. 1984). Sperm motility may be detected by the innervation of the ventral receptacle (Miller 1950) and so as the number of sperm in storage decreases below some threshold, the detection of sperm motility would decrease, causing females to remate (Gromko et al. 1984). Drosophila sperm exhibit two kinetic parameters, minor (beat frequency) and major (coiling diameter) waves (Bressac et al. 1991b). In sperm monomorphic species, the minor and major waves are increased once they are transferred to the female (Bressac et al. 1991b). In sperm heteromorphic D. pseudoobscura, however, only the minor and major waves of long sperm increase within the female (Bressac et al. 1991b). Thus, if obscura group females use sperm motility to detect the number of sperm in storage, then female remating behaviour should be controlled by long, more motile sperm rather than short, less motile sperm. As stated previously, long sperm apparently play no role in female remating behaviour. In conclusion, no relationship was found between the actual number and proportion of long sperm in SSOs and female remating behaviour, and the only significant differences in sperm load between nonvirgin receptive and nonreceptive females was the proportion of each sperm type in the minor SSO, the ventral receptacle. Yet the most robust results in this study associated the probability of female remating with: (1) the absence of an egg in the uterus, and (2) a positive relationship between fecundity and productivity (and thus, the use of fertilizing sperm). The negative sperm load results, in conjunction with the positive progeny production results, suggest oviposition per se, rather than sperm, may be the primary influence determining female remating.
Acknowledgments This research was conducted in the laboratory of T. A. Markow and supported by a National Science Foundation grant DEB-9224263 and a Department of Zoology Graduate Student Research Fund. Special thanks goes to David Capco and Richard Satterlie for equipment use, Matt McGaughey and Matt DeAngelis for data collection, C. R. B. Boake for support during portions of the experimentation and T. L. Karr for support during portions of the writing. T. A. Markow, S. Arnold and J. B. Jones, Jr provided helpful discussion of the data and/or comments on the manuscript.
References Beatty, R. A. & Sidhu, N. S. 1970. Polymegaly of spermatozoan length and its genetic control in Drosophila species. Proceedings of the Royal Society of London, Series B, 71, 14–28. Bellis, M. A., Baker, R. R. & Gage, M. J. G. 1990. Variation in rat ejaculates consistent with the kamikaze-sperm hypothesis. Journal of Mammalogy, 71, 479–480. Birkhead, T. R. & Møller, A. P. 1992. Sperm Competition in Birds: Evolutionary Causes and Consequences. London: Academic Press. Bressac, C. & Hauschteck-Jungen, E. 1996. Drosophila subobscura females preferentially select long sperm for storage and use. Journal of Insect Physiology, 42, 323–328. Bressac, C., Joly, D., Devaux, J. & Lachaise, D. 1991a. Can we predict the mating pattern of Drosophila females from the sperm length distributions in males? Experientia, 47, 111–114. Bressac, C., Joly, D., Devaux, J., Serres, C., Feneux, D. & Lachaise, D. 1991b. Comparative kinetics of short and long sperm in sperm dimorphic Drosophila species. Cell Motility and the Cytoskeleton, 19, 269–274. Briskie, J. V. & Montgomerie, R. 1992. Sperm size and sperm competition in birds. Proceedings of the Royal Society of London, Series B, 247, 89–95. Brooks, D. R. & McLennan, D. H. 1991. Phylogeny, Ecology and Behavior. Chicago: University of Chicago Press. Cobbs, G. 1977. Multiple insemination and male sexual selection in natural populations of Drosophila pseudoobscura. American Naturalist, 111, 641–656. Cook, P. A. & Gage, M. J. G. 1995. Effects of risks of sperm competition on the numbers of eupyrene and apyrene sperm ejaculated by the moth Plodia interpunctella (Lepidoptera: Pyralidae). Behavioral Ecology and Sociobiology, 36, 261–268. Dickinson, J. L. 1986. Prolonged mating in the milkweed leaf beetle Labidomera clivicollis clivicollis (Coleoptera: Chrysomedlidae): a test of the ‘sperm-loading’ hypothesis. Behavioral Ecology and Sociobiology, 18, 331–338. Drummond, B. A. 1984. Multiple mating and sperm competition in the Lepidoptera. In: Sperm Competition and the Evolution of Animal Mating Systems (Ed. by R. L. Smith), pp. 291–370. London: Academic Press. Gage, A. R. & Barnard, C. J. 1996. Male crickets increase sperm number in relation to competition and female size. Behavioral Ecology and Sociobiology, 38, 349–353. Gage, M. J. G. 1991. Risk of sperm competition directly affects ejaculate size in the Mediterranean fruit fly. Animal Behaviour, 42, 1036–1037. Gage, M. J. G. 1994. Associations between body size, mating pattern, testis size and sperm lengths across butterflies. Proceedings of the Royal Society of London, Series B, 258, 247–254. Gage, M. J. G. & Cook, P. A. 1994. Sperm size or numbers? Effects of nutritional stress upon eupyrene and apyrene sperm production strategies in the moth Plodia interpunctella (Lepidoptera: Pyralidae). Functional Ecology, 8, 594–599. Gomendio, M. & Roldan, E. R. S. 1991. Sperm competition influences sperm size in mammals. Proceedings of the Royal Society of London, Series B, 243, 181–185. Gromko, M. H. & Markow, T. A. 1993. Courtship and remating in field populations of Drosophila. Animal Behaviour, 45, 253–262. Gromko, M. H., Newport, M. E. A. & Kortier, M. G. 1984. Sperm dependence of female receptivity to remating in Drosophila melanogaster. Evolution, 38, 1273–1282. Harvey, P. H. & Pagel, M. D. 1991. The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press. He, Y., Tanaka, T. & Miyata, T. 1995. Eupyrene and apyrene sperm and their numerical fluctuations inside the female reproductive tract of the armyworm, Pseudaletia separata. Journal of Insect Physiology, 41, 689–694.
SNOOK: SPERM COMPETITION AND SPERM HETEROMORPHISM 1507
Joly, D. & Lachaise, D. 1994. Polymorphism in the sperm heteromorphic species of the Drosophila obscura group. Journal of Insect Physiology, 40, 933–938. McLain, D. K. 1989. Prolonged copulation as a post-insemination guarding tactic in a natural population of the ragwort seed bug. Animal Behaviour, 38, 659–664. Manning, A. 1962. A sperm factor affecting the receptivity of Drosophila melanogaster females. Nature, 194, 252–253. Miller, A. 1950. The internal anatomy and histology of the imago of Drosophila melanogaster. In: Biology of Drosophila (Ed. by M. Demerec), pp. 420–534. New York: J. Wiley. Parker, G. A. 1970. Sperm competition and its evolutionary consequences in the insects. Biological Review, 45, 525–567. Parker, G. A. & Simmons, L. W. 1994. Evolution of phenotypic optima and copula duration in dungflies. Nature, 370, 53–56. Pasini, M. E., Redi, C. V., Caviglia, O. & Perotti, M. E. 1996. Ultrastructual and cytochemical analysis of sperm dimorphism in Drosophila subobscura. Tissue and Cell, 28, 165–175. Patterson, J. T. 1954. II. The fate of the sperm in the reproductive tract of the Drosophila female in homogamic matings. University of Texas Publication, Genetics, 5422, 19–37. Pitnick, S. 1996. Investment in testes and the cost of making long sperm in Drosophila. American Naturalist, 148, 57–80. Pitnick, S. & Markow, T. A. 1994a. Male gametic strategies: sperm size, testes size, and the allocation of ejaculate among successive mates by the sperm-limited fly Drosophila pachea and its relatives. American Naturalist, 143, 785–819. Pitnick, S. & Markow, T. A. 1994b. Large male advantages associated with costs of sperm production in Drosophila hydei, a species with giant sperm. Proceedings of the National Academy of Sciences, U.S.A., 91, 9277–9281. Pitnick, S., Markow, T. A. & Spicer, G. S. 1995. Delayed male maturity is a cost of producing large sperm in Drosophila. Proceedings of the National Academy of Sciences, U.S.A., 92, 10 614–10 618. Pruzan-Hotchkiss, A., Dejianne, D. & Faro, S. H. 1981. Sperm utilization in once- and twice-mated Drosophila pseudoobscura females. American Naturalist, 118, 37–45. Pyle, D. W. & Gromko, M. H. 1978. Repeated mating by female Drosophila melanogaster: the adaptive importance. Experientia, 34, 449–450. Sanger, W. G. & Miller, D. D. 1973. Spermatozoan length in species of the Drosophila affinis subgroup. American Midland Naturalist, 90, 485–489. Schwartz, J. M. & Boake, C. R. B. 1992. Sexual dimorphism in remating in Hawaiian Drosophila species. Animal Behaviour, 44, 231–238.
Silberglied, R. E., Shepherd, J. G. & Dickinson, J. L. 1984. Eunuchs: the role of apyrene sperm in Lepidoptera? American Naturalist, 123, 255–265. Simmons, L. W., Craig, M., Llorens, T., Schinzig, M. & Hosken, D. 1993. Bushcricket spermatophores vary in accord with sperm competition and parental investment theory. Proceedings of the Royal Society of London, Series B, 251, 183–186. Siva-Jothy, M. T. 1987. Variation in copulation duration and the resultant degree of sperm removal in Orthetrum cancellatum (L.) (Libellulidae: Odonata). Behavioral Ecology and Sociobiology, 20, 147–151. Sivinski, J. 1980. Sexual selection and insect sperm. Florida Entomologist, 63, 99–111. Smith, R. L. 1984. Sperm Competition and the Evolution of Animal Mating Systems. London: Academic Press. Snook, R. R. 1995. The evolution of sperm polymorphism in the Drosophila obscura group. Ph.D. thesis, Arizona State University. Snook, R. R. 1997. Is the production of multiple sperm types adaptive? Evolution, 51, 797–808. Snook, R. R. & Karr, T. L. 1998. Only long sperm are fertilizationcompetent in six sperm-heteromorphic Drosophila species. Current Biology, 8, 291–294. Snook, R. R. & Markow, T. A. 1996. Possible role of nonfertilizing sperm as a nutrient source for female Drosophila pseudoobscura (Diptera: Drosophilidae). Pan-Pacific Entomologist, 72, 121–129. Snook, R. R., Markow, T. A. & Karr, T. L. 1994. Functional nonequivalence of sperm in Drosophila pseudoobscura. Proceedings of the National Academy of Sciences, U.S.A., 91, 11 225–11 229. Thiobold, E. 1979. Stimulation of reproductive activity of females of Acrolepiopsis assectella (Lepidoptera: Hyponomeutoidae) by the presence of eupyrene spermatozoa in the spermatheca. Entomologia Experimentalis et Applicata, 26, 279–290. Tompkins, L. & Hall, J. C. 1983. Identification of brain sites controlling female receptivity in mosaics of Drosophila melanogaster. Genetics, 103, 179–195. Trevitt, S., Fowler, K. & Partridge, L. 1988. An effect of eggdeposition on the subsequent fertility and remating frequency of female Drosophila melanogaster. Journal of Insect Physiology, 34, 821–828. Wilkinson, L. 1990. SYSTAT: the System for Statistics. Evanston, Illinois: SYSTAT. Wedell, N. 1992. Protandry and mate assessment in the wartbiter Decticus verrucivorus (Orthoptera: Tettigoniidae). Behavioral Ecology and Sociobiology, 31, 301–308. Yasui, Y. 1994. Adaptive control of copulation duration by males under sperm competition in the mite, Macrocheles muscaedomesticae. Experimental and Applied Acarology, 18, 543–554.