SEARCH

SEARCH BY CITATION

Keywords:

  • mating behaviour;
  • population differentiation;
  • sepsid flies;
  • speciation;
  • spermathecae;
  • testes

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Theory predicts that males have a limited amount of resources to invest in reproduction, suggesting a trade-off between traits that enhance mate acquisition and those that enhance fertilization success. Here, we investigate the relationship between pre- and post-copulatory investment by comparing the mating behaviour and reproductive morphology of four European and five North American populations of the dung fly Sepsis punctum (Diptera) that display a reversal of sexual size dimorphism (SSD). We show that the geographic reversal in SSD between the continents (male biased in Europe, female biased in North America) is accompanied by differential investment in pre- vs. post-copulatory traits. We find higher remating rates in European populations, where larger males acquire more matings and consequently have evolved relatively larger testes and steeper hyper-allometry with body size. American populations, in sharp contrast, display much reduced, if any, effect of body size on those traits. Instead, North American males demonstrate an increased investment in mate acquisition prior to copulation, with more mounting attempts and a distinctive abdominal courtship display that is completely absent in Europe. When controlling for body size, relative female spermathecal size is similar on both continents, so we find no direct evidence for the co-evolution of male and female internal reproductive morphology. By comparing allopatric populations of the same species that apparently have evolved different mating systems and consequently SSD, we thus indirectly demonstrate differential investment in pre- vs. post-copulatory mechanisms increasing reproductive success.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Understanding how sexual selection contributes to phenotypic divergence within and between species has received considerable interest in evolutionary biology. Sexual selection, as originally conceived by Darwin, describes the variation in reproductive success due to differences among individual males in acquiring mates (Darwin, 1871). However, it is now clear that sexual selection often extends far beyond the initiation of copulation. Post-copulatory processes such as cryptic female choice (Eberhard, 1985, 1996) and sperm competition (Parker, 1970; Simmons, 2001b, 2005) as well as sexual conflict over control of fertilization (Parker, 1979; Arnqvist & Rowe, 2005; Parker, 2006) are recognized as important determinants of reproductive success in polyandrous species that mate multiply. Recent years have witnessed increased research across a broad range of taxa on the diversity of male adaptations that serve to enhance a male's fertilization success relative to that of other males (Simmons, 2001a). However, an issue remains as to how pre- and post-copulatory sexual selection interact in shaping the evolution of trait complexes (Arnqvist & Danielsson, 1999; Simmons, 2001b; Markow, 2002; Emlen et al., 2005a).

Males have limited resources to invest in reproduction, which they must allocate to mate acquisition and, given copulations, successful inseminations and fertilizations (Ball & Parker, 1996; Parker et al., 1997; Simmons, 2001a). This suggests a fundamental trade-off between traits that enhance mating success and those that influence fertilization success, the combination of which is expected to vary with the mating system (Simmons & Emlen, 2006; reviewed in Parker & Pizzari, 2010). For instance, in strongly polyandrous groups with female-biased sex ratio, high female re-mating rates could intensify sperm competition but may relax male competition over access to receptive females at breeding sites. The opposite pattern may be expected if females rarely re-mate implying relaxed sperm competition but perhaps more intense competition among males prior to mating (Reuter et al., 2008). Examples of an allocation trade-off are apparent in the evolutionary diversification of beetle horns (Emlen et al., 2005b; Simmons et al., 2007), the contrasting patterns of courtship display traits and sperm characteristics among lineages of Drosophila (Pitnick, 1996; Markow, 2002), or the association between male courtship attractiveness and paternity share in fireflies (Demary & Lewis, 2007).

While certain traits will be predominantly favoured by either pre- or post-copulatory sexual selection, other traits, such as body size, are clearly important in both (Holleley et al., 2006). Large body size, as well as the expression of countless and diverse primary and secondary sexual traits, is often favoured by classic male–male competition or female choice (reviewed in Andersson, 1994; Blanckenhorn, 2000; Fairbairn et al., 2007; Hunt et al., 2009). For instance, during aggressive contests for access to females that are typical in many animal species, large males are often more successful at acquiring mates and/or forcing copulations (Parker & Thompson, 1980; Zucker & Murray, 1996; Shine & Mason, 2005; Brown, 2008; Hasegawa et al., 2011; Jorge & Lomonaco, 2011). Large body size can also confer a mating advantage in noncombative courtship displays, as observed in certain insects and anurans (Howard & Young, 1998; Simmons, 1988). The effect of body size on mate acquisition is particularly apparent in species that display size-dependent alternative mating tactics. In a number of fishes and birds (Ryan et al., 1992; Brantley & Bass, 1994; Lank et al., 1995) and even insects such as the rove beetles and yellow dung flies (Forsyth & Alcock, 1990; Pitnick et al., 2009), large and small males can develop completely different precopulatory strategies towards attaining matings that might involve sneaking, cuckoldry or even female mimicry (reviewed in Gross, 1996; Schuster & Wade, 2003).

Body size also affects post-copulatory processes via correlated morphology and allometric scaling (Simmons, 2001a). Relative testis size, that is, sperm production in relation to body size, is commonly used to gauge sperm competitive ability. In general, testis size scales positively, albeit typically hypo-allometrically with body size, both within (Gage et al., 1995; Tomkins & Simmons, 2002; Wedell et al., 2006) and across species (Møller, 1988, 1989; Gage, 1994; Hosken, 1997; Stockley et al., 1997; Schulte-Hostedde & Millar, 2004; Minder et al., 2005; Schulte-Hostedde & Alarie, 2006; Liao et al., 2011; Vahed et al., 2011). Because larger males consequently harbour absolutely more but relatively fewer sperm, they are presumed to transfer or displace larger quantities of sperm in many species, conferring a fertilization advantage (reviewed in Simmons, 2001a; Bangham et al., 2002). In other species, however, there is no size advantage (Stockley & Purvis, 1993; Parker & Simmons, 1994; Arnqvist & Danielsson, 1999; Tomkins & Simmons, 2002) or at times even a small male advantage in sperm competition (Schneider et al., 2000; Danielsson, 2001; Sato et al., 2004; Schneider & Elgar, 2005; Wenninger & Averill, 2006; Watt et al., 2011), such as when smaller males invest disproportionally more in testes or ejaculates (Schulte-Hostedde & Millar, 2004; Schäfer et al., 2008; Schutz et al., 2010).

We here study the relationship between pre- and post-copulatory sexual selection by comparing the mating behaviour and reproductive morphology of four European and five North American populations of the dung fly Sepsis punctum (Fabricius, 1794) (Diptera: Sepsidae). Sepsid flies are increasingly used as model organisms in sexual selection studies because they have diverse sexual dimorphisms as well as elaborate mating behaviour (Puniamoorthy et al., 2008, 2009) and can be reared easily in the laboratory (e.g. Teuschl & Blanckenhorn, 2007). Sepsis punctum is a geographically widespread species that can be collected not only on cattle pastures but also on dog excrements in parks and open fields (Pont & Meier, 2002). In a recent study (Puniamoorthy et al., 2012), we found that the intensity of precopulatory sexual selection acting on male body size was much stronger in European than in North American populations. In agreement with the differential equilibrium hypothesis of sexual size dimorphism (SSD) (Blanckenhorn, 2000; Fairbairn et al., 2007), this can explain the geographic reversal in SSD of S. punctum observed between the continents. Schultz (1999) first noticed that the presence of precopulatory courtship also varies between the continents, suggesting that the reversal in SSD might segregate with differences in the mating system and other trait complexes. We here explore this further by focusing on traits related to mate acquisition and traits with putative function in sperm competition. We took an integrated approach, comparing detailed behavioural experiments and observations with morphological measures of fertilization-related structures and body size across nine cross-continental S. punctum populations.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Sampling of populations

We collected flies from four European sites, Nyköping, Sweden (58.67° N, 16.94° E), Berlin, Germany (52.45° N, 13.28° E), Vienna, Austria (48.20° N, 16.36° E), Zürich, Switzerland (47.40° N, 8.55° E) as well as five North American populations from Davis, California (38.54° N, −121.75° W), Park City, Utah (40.66° N, −111.52° E), Athens, Georgia (33.96° N, −83.38° E), Manhattan, New York (40.78° N, −73.96° E) and Ottawa, Ontario (45.42° N, −75.67° E). We caught gravid females on and around fresh dung pats in open cow pastures, transported them back to the laboratory in Zurich, and used them to establish stock cultures of 10–20 iso-female lines per population. Alternatively, we set out small pots of cow dung in city parks overnight for a few days and shipped them back to the laboratory. The emergent flies from each pot were treated as single lines. All fly cultures were housed in separate clear plastic containers, reared in a climate chamber at standardized 24 °C, 60% humidity, 14-h light cycle and were regularly supplied with fresh cow dung, sugar and water ad libitum.

Rearing of flies for experiments

To generate a range of phenotypic body sizes, we provided stock cultures of each population pots with different amounts of cow dung and allowed for oviposition overnight. We transferred these pots into another container and reared them under the above-mentioned standard conditions. After approximately 2 weeks of juvenile development, we sexed emerging flies individually under a microscope within 24 h of eclosion and subsequently housed virgin males and females in separate containers.

Morphological study

For each population, we randomly selected approximately 50–140 individuals from the ‘virgin’ containers and froze them at −20 °C overnight. We then measured the flies for body size (head width) before dissecting them under the microscope in Ringer's solution under a Leica MS 5 microscope (Leica Microsystems, Wetzlar, Germany). We transferred both male testes and both female spermathecae on a concave glass slide with a drop of Ringer's and cover slide. We calculated the volume of the respective reproductive structures from the measurements of the length and width of both testes (ellipsoid) and the diameter of both spermathecae (sphere) under a Zeiss light microscope (Carl Zeiss Group, Oberkochen, Germany). We had assistance from several students (blocking factor in the statistical analysis).

Behavioural study

Mating trials with virgin flies

We conducted all mating trials 3–4 days after eclosion to ensure sexual maturity (cf. Teuschl & Blanckenhorn, 2007; Puniamoorthy et al., 2012). We randomly selected a male and a female from the ‘virgin’ containers and introduced them into a clear glass vial (containing cow dung smeared on a small filter paper) to observe their interaction for a maximum of 1 h or until copulation occurred. We recorded the number of male mounting attempts, the number of courtship displays, as well as the copulation duration (cf. Ding & Blanckenhorn, 2002). We conducted these mating trials until we reached our targeted sample size of approximately 20 mated pairs per population.

Re-mating trials with mated individuals

Upon successful copulation, we separated the mated pair, housed each male and female in a new glass vial (with dung, sugar and water) and gave them individual identification labels. One week after the first copulation, we conducted re-mating trials, randomly assigning each male to a new female and again recorded all interactions for a maximum of 1 h in a new glass chamber. At the end of the trial, we returned each fly to its individual ‘home’ vial (replenished with fresh dung, sugar and water). We repeated these re-mating trials for both sexes for a maximum of 8 weeks or until the flies died, at which point they were frozen and measured for body size.

Statistical analyses

To quantify differential allocation in male precopulatory courtship display, we created an index by summing both the number of male mounting attempts and the number of courtship displays (both counts), correcting for the duration of each trial (via residuals). We scored copula duration, mating and re-mating frequencies as separate dependent variables. For the reproductive structures, we took the mean (of two) testes volume and mean spermathecal volume.

We performed significance testing using ancova with continent as fixed, population nested within continent as random factor and body size as a continuous covariate (unless otherwise mentioned). Body size was z-score standardized before analysis in all ancovas such that all factors, and especially the main effects, are properly evaluated at the centre of the actual data distribution. For overall body size, copulation duration and re-mating frequencies, sex was included as an additional factor. All volume measurements were cube-root-transformed to the linear scale. We initially included all relevant interaction terms, which were dropped from the model if not significant, except when required as error terms in testing higher level effects (for details, see Data S1). We conducted all analyses using the software IBM SPSS Statistics version 19.0 (SPSS, Inc., IBM Corporation, Armonk, NY, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Scaling relationships between traits

As expected by our earlier study (Puniamoorthy et al., 2012), we found SSD to be female biased in North America and male biased in Europe (continent by sex interaction for body size: F1,7 = 20.23, P = 0.002; Table 1). Populations within continents also varied with respect to overall body size (F7,879 = 6.71, P = 0.008), and there was additional (but uninteresting) variation introduced by measurers (blocking effect: F18,879 = 3.91, P < 0.001).

Table 1. Population means (±SE) for morphology data: body size, average male testes volume, and average female spermathecal diameter and volume [n, sample size]
PopulationFemaleMale
Ave. spermathecal diameter (×10 mm)Ave. spermathecal volume (×10−3 mm3)Head width (mm) n Ave. testes volume (×10−2 mm3)Head width (mm) n
Austria0.919 ± 0.0950.220 ± 0.0620.976 ± 0.109230.718 ± 0.4691.043 ± 0.18027
Germany0.959 ± 0.1000.254 ± 0.0920.970 ± 0.119590.568 ± 0.3500.977 ± 0.12159
Switzerland0.958 ± 0.1230.250 ± 0.1030.873 ± 0.084880.569 ± 0.4430.934 ± 0.13147
Sweden0.930 ± 0.0990.228 ± 0.0760.884 ± 0.089370.523 ± 0.2800.938 ± 0.10073
California0.845 ± 0.0940.173 ± 0.0560.833 ± 0.105670.163 ± 0.0660.871 ± 0.07657
Georgia0.924 ± 0.1000.224 ± 0.0720.907 ± 0.068730.200 ± 0.0610.898 ± 0.08152
New York0.906 ± 0.0440.206 ± 0.0290.986 ± 0.030250.149 ± 0.0460.947 ± 0.05421
Ottawa0.947 ± 0.0820.237 ± 0.0670.929 ± 0.040750.181 ± 0.0560.912 ± 0.05757
Park city0.937 ± 0.0720.227 ± 0.0470.890 ± 0.088190.214 ± 0.0450.990 ± 0.02821

Female spermathecal size was strongly positively (but hypo-allometrically) related to body size (Fig. 1a; F1,439 = 141.07, P < 0.001; overall regression equations based on log-transformed linear measures (±95% CI): [Europe] = −1.281 (±0.069) + 0.437x (±0.116), [N. America] = −1.389 (±0.077) + 0.585x (±0.131)). The average spermathecal volume ranged from 0.173 ± 0.056 SE (×10−3 mm3) in the California population to 0.254 ± 0.092 SE (×10−3 mm3) in the German population (Table 1). There was variation due to measurer and between populations (blocking effect: F17,439 = 5.76, P < 0.001; population effect: F7,439 = 9.27, P < 0.001). However, relative (i.e. size-controlled) spermathecal size did not vary significantly among continents (continent effect: F1,7 = 0.876, P = 0.378). Furthermore, the allometric relationship between spermathecae and body size was the same for all populations and on both continents (i.e. population by body size and continent by body size interactions were n.s.; for details see Data S1).

image

Figure 1. (a) Cube-root-transformed (linearized) spermathecae volume against female body size, and (b) cube-root-transformed testis volume against male body size.

Download figure to PowerPoint

Overall, we found that larger males had bigger testes (strong main effect of body size: F1,378 = 623.27, P < 0.001). The average testes volume varied drastically among populations up to five-fold, ranging from 0.149 ± 0.046 SE (×10−2 mm3) in New York to 0.718 ± 0.469 SE (×10−2 mm3) in Austria. Even after controlling for body size, both the relative testes size and the testes–body size allometry within continents were significantly different (population main effect: F7,378 = 7.86, P < 0.001; population by body size interaction: F7,378 = 10.13, P < 0.001). Crucially, there was an especially strong systematic difference between the continents (continent main effect: F1,7 = 102.83, P < 0.001), with the populations in Europe having larger testes and testes displaying a much steeper hyper-allometric relationship with body size (Fig. 1b; body size by continent interaction: F1,7 = 24.77, P = 0.002; regression equations based on log-transformed linearized measures (±95% CI): [Europe] = −1.285 (±0.069) + 1.389x (±0.114), [N. America] = −0.911 (±0.102) + 0.551x (±0.174)).

Mounting attempts

The number of male mounting attempts until successful copulation differed significantly between the continents (Table 2). Many copulations in Europe were attained by the first male mounting attempt, whilst American males had to work much harder to gain a successful mating, ranging from 2.70 ± 2.32 SE attempts in California to 5.14 ± 2.19 SE attempts in Ottawa (Fig. 2a; continent main effect: F1,7 = 12.56, P = 0.009; population main effect: F7,163 = 7.32, P < 0.001; body size effect and all body size by factor interactions n.s.).

image

Figure 2. Average (±SE) (a) mounting attempts; (b) precopulatory investment, that is, combined mounting and courtship displays. Dashed circle highlights American west-coast populations.

Download figure to PowerPoint

Table 2. Population means (±SE) for body size and male behavior data: pre- and post-copulatory traits [‘/’ means no data available; n, sample size]
PopulationMale head width (mm) n Investment in mate acquisitionInvestment in post-copulatory selection
Ave. no. of mountsAve. no. of courtship displaysAve. precop. investment (corrected for duration)Ave. no. of re-mating trialsAve. no. of copulationsAve. duration of first copulation (min)Ave. re-mating frequency (corrected for number of trials)
Austria1.04 ± 0.15271.41 ± 0.9701.41 ± 0.915.67 ± 2.861.78 ± 0.9320.37 ± 6.111.77 ± 0.91
Germany1.01 ± 0.05211.62 ± 1.4001.62 ± 0.986.62 ± 1.502.14 ± 1.0623.11 ± 3.711.96 ± 0.98
Switzerland1.12 ± 0.05242.04 ± 1.0402.04 ± 0.865.17 ± 2.531.62 ± 0.8223.96 ± 5.821.63 ± 0.65
Sweden1.09 ± 0.14231.35 ± 0.7101.35 ± 0.656.74 ± 2.012.22 ± 0.9021.74 ± 5.642.22 ± 0.81
California0.91 ± 0.07202.70 ± 2.321.20 ± 1.853.90 ± 2.126.10 ± 2.201.20 ± 0.4128.11 ± 8.481.19 ± 0.40
Georgia0.93 ± 0.10182.39 ± 2.0012.61 ± 14.2815.00 ± 8.514.17 ± 1.761.06 ± 0.2423.17 ± 5.111.06 ± 0.22
New York0.87 ± 0.10213.05 ± 2.0016.23 ± 15.5119.27 ± 7.154.86 ± 2.681.05 ± 0.2122.92 ± 2.361.05 ± 0.22
Ottawa0.90 ± 0.0575.14 ± 2.1914.71 ± 11.7319.86 ± 5.48//24.84 ± 4.37/
Park city0.89 ± 0.09124.50 ± 4.250.83 ± 1.405.33 ± 2.01//22.80 ± 2.84/

Precopulatory courtship display and intensity

One major difference between the two continents is the absence of an abdominal courtship display in all four European populations (Table 2; see Movie S1). Additionally, the intensity of displays varied strongly among the American populations, with relatively low occurrence in the California and Park City populations. Consequently we found significant variation between the continents as well as among the populations for the combined index of investment in mate acquisition (subsuming mounting and courtship attempts), the latter being driven mainly by the variation between the courting American populations (Fig. 2b; continent effect: F1,7 = 9.98, P = 0.039; population effect: F7,163 = 30.84, P < 0.001; body size main effect and all body size by factor interactions n.s.).

Female mating rates

The mating rate of virgin flies varied strongly between populations (matings/total trials: Austria = 27/50; Germany = 25/49; Switzerland = 24/70; Sweden = 23/57; California = 20/93; Georgia = 18/109; New York = 22/88; Ottawa = 7/170; Park City = 12/130), with the European populations mating more readily than the American populations (binary logistic model on 1/0 data; continent effect: inline image = 66.50, P < 0.001; population effect: inline image = 38.16, P < 0.001).

Male and female re-mating behaviour

Due to the extremely low mating rates in the Ottawa and Park City populations, we did not include them in the re-mating study. For the remaining seven populations, we conducted weekly mating trials and found that the number of copulations over 6–8 weeks varied between the continents (Table 2; continent main effect: F1,280 = 27.02, P = 0.003). Interestingly, larger females tended to re-mate less frequently, whilst larger males attained more copulations (Fig. 3b). This effect was exclusively driven by the European populations (sex by body size interaction: F1,280 = 12.53, P < 0.001; three-way sex by continent by body size interaction: F2,10 = 6.67, P = 0.014; body size main effect and all other factor by body size interactions n.s.)

image

Figure 3. (a) Copula duration in females and males in nine populations, and (b) re-mating frequency in seven populations as a function of body size.

Download figure to PowerPoint

Copulation duration

Copulation duration in S. punctum typically varied from 20 to 30 min (Table 2), strongly depending on body size such that larger females and smaller males copulated for longer (Fig. 3a; male body size effect: F1,162 = 43.93, P < 0.001; female body size effect: F1,162 = 10.08, P = 0.002). Despite differences between populations, the continental origin of the flies did not significantly influence copulation duration (continent main effect: F1,7 = 1.95, P = 0.205; population main effect: F7,162 = 3.32, P = 0.002; all corresponding factor by body size interactions n.s.).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

We show that the geographic reversal in SSD between European and North American populations of the dung fly S. punctum is accompanied by differential allocation in traits engaged in pre- vs. post-copulatory sexual selection. European populations display a higher mating propensity and males have evolved relatively larger testes and much steeper, positive or hyper-allometry with body size, in accordance with sperm competition theory predicting higher investment in sperm production with increasing level of sperm competition (Parker et al., 1997; Parker & Pizzari, 2010). In sharp contrast, North American populations show a much lower female mating rate, and males invest more in mate acquisition prior to copulation. Their mating system is characterized by more male mounting attempts and by the presence of a distinctive abdominal courtship display, which is completely absent in Europe. At the same time, we found the intensity of precopulatory sexual selection on male body size, in terms of the cumulative number of matings over a significant portion of the lifetime (6–8 weeks), to be much stronger in European populations displaying male-biased SSD than in North American populations with female-biased SSD (Fig. 3b), corroborating previous findings based on single mating probabilities (Puniamoorthy et al., 2012). By comparing allopatric populations of the same species that apparently have evolved different mating systems and consequently SSD, we thus demonstrate differential allocation of European and North American flies in pre- vs. post-copulatory mechanisms affecting reproductive success (Markow, 2002), analogous to comparisons among intra-specific morphs with alternative mating strategies (Gage et al., 1995; Tomkins & Simmons, 2002; Kelly, 2008).

Theory predicts an optimal mating rate for females beyond which multiple matings can have detrimental effects (Arnqvist, 1989; Firman & Simmons, 2008; Simmons & Garcia-Gonzalez, 2008), as copulation can increase predation risk, decrease foraging ability (Daly, 1978; Sih et al., 1990; Fairbairn, 1993), or even produce internal injury due to male genital structures (Crudgington & Siva-Jothy, 2000; Blanckenhorn et al., 2002). Given that males usually profit from multiple matings and females often do so to a much lesser extent, this can potentially generate conflict over mating and fertilization (Bateman, 1948; Arnqvist et al., 2000; Arnqvist & Rowe, 2005). Studies show that in species with male-biased SSD, male resource defence polygyny, territoriality, and monopolization of females are common, whereas in species with female-biased (or no) SSD, female choice and male courtship displays dominate (Andersson, 1994; Ding & Blanckenhorn, 2002; Arnqvist & Rowe, 2005). This predicts that external or internal courtship facilitating female choice should be more apparent, or should more likely evolve, in species (or populations) with female-biased SSD (Eberhard, 1996), as is the case for North American but not European S. punctum, the latter displaying male-biased SSD.

Moreover, according to theory (Parker et al., 1997), female re-mating rate should positively correlate with relative testis size both within (e.g. Gage, 1994; Gage et al., 1995; Wedell, 1997; Tomkins & Simmons, 2002; this study) and among species (e.g. Møller, 1988, 1989; Gage, 1994; Hosken, 1997; Stockley et al., 1997; Minder et al., 2005). Thus European S. punctum males, which face greater risk of sperm competition because of higher female (re-)mating rates, evolved relatively larger testes and a steep testes/body size hyper-allometry as compared to North American populations of this species, presumably in connection with the change in mating system (Fig. 1b). Although it could in principle relate to different sperm competition mechanisms, this difference in allometry most parsimoniously reflects the necessity of producing more sperm to increase fertilization chances in a fair or loaded raffle (Parker, 1970; Short, 1979; Simmons, 2001a).

Copulation duration and male investment in transferred ejaculates can depend on various factors such as the risk of sperm competition and/or the quality of the females (reviewed in Kelly & Jennions, 2011). In fact, a recent study of the closely related Sepsis cynipsea documented that males invest more sperm in more fecund females with smaller males copulating longer and copulations lasting longer with larger females (Teuschl et al., 2010; cf. Lefranc & Bundgaard, 2000; for Drosophila). Our data (Fig. 3a) similarly indicate that, equally on both continents, copulation duration in S. punctum increases with female size and decreases with male size, presumably because, physiologically, larger females require and can store more sperm because they produce more eggs, and probably larger males have wider ducts and can transfer more sperm quickly (Simmons, 2001b; Blanckenhorn et al., 2004). An additional or alternative explanation could be a potential trade-off between polyandry and relative ejaculate expenditure (Vahed & Parker, 2012). For instance, given their increased mating rate, larger males might copulate for shorter time because they invest less sperm per copulation, and, conversely, small males invest relatively more per copulation because of their reduced future mating probability (Parker & Ball, 2005; Fromhage et al., 2008; Vahed et al., 2011). So far, little is known about the detailed mechanisms of sperm competition in any sepsid species. Based on behavioural observations and data on male sperm investment, Martin & Hosken (2002) concluded a ‘raffle’ mechanism of sperm competition for the related S. cynipsea. However, preliminary paternity analysis in S. punctum based on twelve doubly mated females of a German population indicates almost complete last male sperm precedence (Schultz, 1999), so sperm competition mechanisms might be quite variable among Sepsis species. Clearly, the relationship between ejaculate allocation and relative paternity success in sepsid flies requires further study.

Sexual selection has been considered as an important barrier to gene flow through its direct effects on mate or gamete recognition (Panhuis et al., 2001; Servedio & Noor, 2003; Coyne & Orr, 2004). Several comparative studies have documented co-evolution between traits engaged in insemination and fertilization (including testis size and female reproductive tract morphology) (Presgraves et al., 1999; Brown & Eady, 2001; Minder et al., 2005; Rugman-Jones & Eady, 2008; Rönn et al., 2011; Thüler et al., 2011), but this appears not to be the case in S. punctum. In contrast to testes size, which was highly differentiated between American and European populations, the number, relative size and allometry of the female spermathecae (the sperm storing organs) is highly conserved across the continents and populations, consistent with a recent comparative study of the internal female reproductive tract of 41 sepsid species (Puniamoorthy et al., 2010). Said study additionally documents some morphological variation in a secondary sperm storage organ, the ventral receptacle. In other acalyptrate Diptera, such as the Mediterranean fruit fly, the ventral receptacle is the first sperm storage organ to deplete, suggesting that these delicate structures could be the likely site of fertilization (De Carlo et al., 1994). However, we did not treat this in our study because in S. punctum this unsclerotized, membranous structure, which is very much smaller than the spermathecae (approximately < 0.1 mm), is difficult to visualize without staining and even harder to dissect without destroying it.

We document that male courtship behaviour varied significantly both between and within continents. In particular, our study suggests that among the North American populations, there could be a strong east–west gradient in the intensity of male display (Fig. 2b). Furthermore preliminary data based on neutral molecular markers indicate genetic differentiation within American populations of S. punctum (N. Puniamoorthy, unpublished data). This scenario is consistent with mountain ranges and glaciation events in America that have been documented to limit gene flow and result in population divergence and speciation (e.g. Hewitt, 2004; Mirol et al., 2007). Further work on variation in courtship in North America is needed to investigate the significance of the abdominal display in possibly establishing precopulatory barriers to gene flow.

In summary, we demonstrate a shift in mating system and associated changes in behaviour and morphology when comparing cross-continental populations of the widespread dung fly S. punctum. North American populations of this species are smaller, display female-biased SSD, low (re-)mating rates, high investment in precopulatory courtship and mating attempts, and hypo-allometric testes/body size scaling, whereas European populations are larger in size, display male-biased SSD, higher (re-)mating rates, no precopulatory courtship, and steeper hyper-allometric testes/body size scaling (Figs 1 and 2). Because the allometry of female spermatheca size to body size is similar on both continents, we have no indication of co-evolution of male and female internal reproductive morphology by sexual selection or conflict (cf. Thüler et al., 2011). Our results imply and demonstrate, across populations of a single species, differential investment in pre- vs. post-copulatory traits indicative of a putative trade-off (Parker et al., 1997; Markow, 2002). Which mating system or SSD is the original state, and which sequence of events lead to the divergent evolution of North American and European S. punctum, remains to be answered by further studies.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

We thank members of the Blanckenhorn group for numerous helpful discussions throughout this project. We especially thank D. Berger, R. Walters and C. Dmitriew for sampling some of the S. punctum populations. We also thank Dr Rudolf Meier and the evolution lab in Singapore for the supplementary movie file. In addition, we thank the two anonymous reviewers for constructive and helpful comments on earlier versions of this manuscript. We appreciate all the assistance in the laboratory from U. Briegel, S. Matic and many Bio361 student helpers. All authors have read and contributed the manuscript and declare no conflict of interest. N. Puniamoorthy was supported by the National University of Singapore Overseas Graduate Scholarship, M.A. Schäfer by a grant from the German Research Council, and W.U. Blanckenhorn by a grant from the Swiss National Fund.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
jeb2605-sup-0001-DataS1.pdfapplication/PDF42KData S1 Details of statistical models used in study.
jeb2605-sup-0002-MovieS1.wmvvideo/wmv4263KMovie S1 Pre-copulatory courtship in North American S. punctum (New York).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.