Supplementary Material
Adaptations to sexual selection and sexual conflict: insights from experimental evolution and artificial selection.
Dominic A. Edward, Claudia Fricke and Tracey Chapman
SUPPLEMENTARY MATERIALS AND METHODS
(a) Stocks and cultures
Wild-type: The Dahomey wild-type stock used as the ancestral source for the sex ratio selection lines was collected in 1970 and has been maintained since then at high population sizes in four population cages with overlapping generations. Stock maintenance is as follows: each stock cage is supplied with three bottles (189 ml each) containing 70 ml of sugar-yeast (SY) food (100g brewer’s yeast, 100g sugar, 20g agar, 30mL Nipagin (10% w/v solution), 3 ml Propionic acid, in 1 L water) (Bass et al. 2007) every week. Bottles are removed after 28 days. All population cages, selection lines and experiments were maintained at 25˚C in non-humidified rooms on a 12:12h light:dark cycle.
Marker flies: Marker flies homozygous for a recessive sepia eye colour allele were used to confirm paternity in sperm competition assays. Flies used were from a stock recently backcrossed for four generations into the Dahomey genetic background. sepia stocks were maintained in replicate bottle cultures at large populations sizes.
Sex ratio lines: These lines have been described previously (Wigby & Chapman 2004). The intensity of sexual selection and sexual conflict were both varied by altering the adult sex ratio. Each generation three replicate lines of male-biased (MB, 75 males and 25 females), equal sex (ES, 50 males and 50 females) and female-biased FB (FB, 25 males and 75 females) sex ratio treatments were propagated. The sex ratio in the MB lines was changed to 70 males and 30 females from generation 53 to ease propagation by increasing the number of females. Flies were provided access to water and were fed ad libitum with two vials of SY food with added live yeast every two or three days. Nine days after the cages were set up, eggs were collected. The majority of eggs were allowed to hatch before larvae were collected, thus minimizing selection on early egg hatchability. Larvae were raised at standard density to minimize environmentally determined differences in adult body size. All adults were allowed to eclose over two days before sorting them into their appropriate sex ratio treatments. The lines were selected for about 70 generations until 2007, and then maintained at 18˚C (to slow down generation times by a factor of 2) until March 2008 when a further 26 generations of further selection were applied.
(b) Experimental Procedures
Responses to rivals in FB and MB selection line males
Eggs were collected by allowing females to oviposit upon petri dishes filled with a grape juice medium (50g agar, 600ml red grape juice, 42.5ml Nipagin (10% w/v solution), 1.1l water) smeared with live yeast paste. Larvae were collected within 24hrs following oviposition and reared at a standard density of 100 per vial in SY vials. At eclosion flies were collected and sexes separated using ice anesthesia. Virgin sepia females and sepia males were kept at a density of 10 flies per vial prior to use. Males from each of the MB and FB cultures were collected and placed either alone or in groups of 4 for 5 days before use. SY food medium with ad libitum live yeast granules added was used throughout. On the day before matings, sepia females were placed singly in vials. The following day, males from the sex ratio lines were introduced singly to a female and the latency to mating and mating duration were recorded. After mating, males were removed and discarded. Females were then given an opportunity to remate with a sepia male 24h later. Latency to mating and mating duration were again recorded. For the period between first and second matings the number of eggs and viability of eggs produced by each female was recorded. For the four days after the second mating all offspring were allowed to develop and scored for paternity. Females were transferred daily to new vials under light CO2 anaesthesia. The experiment comprised three replicates each of MB and FB lines all tested in the presence and absence of rivals, to give a total of 12 treatments. Sample sizes ranged from 11 - 38 first matings per treatment (median = 34) and from 4 to 20 rematings per treatment (median = 12).
Single mating fecundity and remating receptivity of FB, ES and MB selection line females following single matings
To test whether females from the selection lines had altered in their responses to the effects of single matings, we also measured the egg production and remating receptivity of FB, ES and MB females following single matings with control males or males lacking male ejaculate sex peptide (SP, Liu & Kubli 2003). SP is known to stimulate egg laying and decrease female receptivity following its transfer during mating. The aim here was to measure whether FB, ES and MB females differed in their responses both in the presence and the absence of SP. SP null (SP0) males were the offspring of delta130 / TM3 sb ry females (the delta 130 deletion covering the SP locus) and SP0 0325/ TM3 Sb ry males (with the SP0 0325 being a mutation in the SP locus). The resulting SP0 0325/ delta 130 male offspring lacked sex peptide. Control (SP+) males were the offspring of delta 130 / TM3 Sb ry females mated to SP+ 0416 / TM3 Sb ry males (with SP+ 0416 having both a wild-type and null version of the SP gene). All stocks were as described in Liu & Kubli (2003).
FB, ES and MB virgin females were collected from standard density cultures as described above and stored 5 per vial. SP+ and SP0 males were collected from relaxed density cultures and stored 2 per vial following eclosion. For the mating tests, 3-4 day old FB, ES and MB virgin females were then placed individually into vials with two virgin SP+ or SP0 males. We recorded the timing of introduction and the start and end of mating. All females that mated were then transferred to a new oviposition vial containing a small blob of live yeast paste for the following 16 hours. The next morning females were transferred into vials each containing two Dahomey wild-type males, and we recorded the time to initiate remating and the number of FB and MB females that remated in the remating test. The whole experiment was conducted over four blocks with 15 replicates for each line in each block, giving a starting sample size of 60 per treatment. The experiments were conducted after generation 70.
(c) Statistical analysis
All covariates were checked for normality and transformed as necessary to satisfy a normal distribution. Latency to mate and mating duration were both transformed by natural logarithm. Replicate vials were discounted in which no viable offspring were produced following the first mating, indicating infertility of the focal male, or no sepia offspring produced following remating, implying infertility of the sepia male. The latency to mate, duration of mating, number of eggs produced in the intermating interval and fertility of eggs produced in the intermating interval were compared using generalized linear mixed effect models. The sex ratio bias of the line (MB/FB), level of sperm competition intensity (+/- rivals) plus the interaction were considered as fixed effects. Line, nested within sex ratio bias, was included as a random effect. Normal and binomial model distributions were used as appropriate. The competitive reproductive success of each focal male, measured as the proportion of offspring following remating, was compared in a generalized linear mixed effects model with a binomial distribution. The sex ratio bias of the line (MB/FB), level of sperm competition intensity (+/- rivals), number of eggs produced in the intermating interval and all two-way interactions were considered as fixed effects. Line, nested within sex ratio bias, was included as a random effect. Models were constructed to initially include all main effects plus interactions and subsequently simplified by stepwise removal of the least significant factor. At each step the current and reduced models were compared in an analysis of variance using an F test or c2 test as appropriate. Factors were removed only where the change did not significantly influence the fit of the model at p=0.05 (Crawley 2005). The responses of females from the MB, ES and FB lines to single mating data were analysed using generalized linear models (GLIMS) and binomial or quasi-Poisson error structures (Crawley 2005). All analyses were conducted in R v2.10.0 (R Foundation for Statistical Computing, Vienna, Austria).
SUPPLEMENTARY MATERIAL RESULTS
FB, ES and MB female receptivity and fecundity following single matings
Fecundity was variable, with significant differences across blocks and between replicates within regimes (table S1a). However, over and above this, there were significant differences between selection regimes in fecundity, due to higher fecundity with ES females rather than to differences between FB and MB females (table S1a). Female fecundity was significantly higher, as expected based on the known phenotype of SP (Liu & Kubli 2003), following matings to SP+ in comparison to SP0 males across all selection regimes. The interaction between selection regime and response to receipt of SP was just significant, with a slightly more marked effect of SP in FB females. In terms of female remating receptivity and intermating interval (table S1b,c), there were also significant differences between replicates within regimes and between blocks. Again as expected, there were significant differences in the receptivity and intermating interval of females of all regimes following matings with either SP+ or SP0 males. There was no obvious pattern however for either trait of consistent differences due to selection regime. Overall, there was no evidence that FB and MB females differed in their receptivity and fecundity following single matings, or that they responded in a qualitatively different way following receipt of SP.
Table S1. FB, ES and MB female fecundity and receptivity following matings to SP+ and SP0 males.
(a) Female fecundity (mean eggs ± se) in a 16h period following single matings to SP+ and SP0 males, the combined means across blocks and replicates are given for illustration. Fecundity was analysed using a GLIM with quasipoisson errors (dispersion parameter = 9.09).
Selection treatment Mean fecundity (± se)
SP+ SP0
FB 24.7 (0.87) 14.0 (0.82)
ES 31.7 (1.34) 23.0 (1.31)
MB 25.8 (1.01) 18.3 (1.15)
Source df Deviance F p
MB/ES/FB treatment 2 489.8 26.74 <0.001
Male genotype 1 903.1 98.61 <0.001
Male genotype x treatment 2 63.8 3.51 0.03
Replicate nested within treatment 6 104 1.91 0.077
Block 3 94.7 3.45 0.016
(b) Female remating propensity following single matings to SP+ and SP0 males, combined means shown for illustration. Remating was analysed using a GLIM with binomial errors.
Selection treatment Mean proportion remating
SP+ SP0
FB 0.60 0.98
ES 0.60 0.94
MB 0.52 0.91
Source df Deviance (c2) p
MB/ES/FB treatment 2 6.94 0.03
Male genotype 1 210.79 <0.001
Male genotype x treatment 2 3.49 0.17
Replicate nested within treatment 6 32.27 <0.001
Block 3 8.31 0.04
(c) Remating interval (mean mins ± se) following single matings to SP+ and SP0 males; combined means across blocks and replicates are given for illustration. The remating interval (time taken from start of the first mating to the start of the second mating) was analysed using a GLIM with quasipoisson errors (dispersion parameter = 4.24).
Selection treatment Mean remating interval (± se)
SP+ SP0
FB 1455 (7.38) 1399 (5.71)
ES 1442 (7.15) 1383 (7.29)
MB 1440 (8.33) 1399 (6.16)
Source df Deviance F p
MB/ES/FB treatment 2 21.6 2.51 0.08
Male genotype 1 352.5 81.63 <0.001
Male genotype x treatment 2 6.8 0.79 0.45
Replicate nested within treatment 6 88.2 3.47 0.002
Block 3 90.3 6.97 <0.001
REFERENCES
Bass, T. M., Grandison, R. C., Wong, R., Martinez, P., Partridge, L. & Piper, M. D. W. 2007 Optimization of dietary restriction protocols in Drosophila. J. Gerontol. A Biol. Sci. Med. Sci. 62, 1071-1081.
Crawley, M. J. 2005 Statistics: An introduction using R. Chichester, West Sussex, UK: John Wiley & Sons, Ltd.
Liu, H., & Kubli, E. 2003 Sex peptide is the molecular basis of the sperm effect in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 100, 9929-9933.
Wigby, S. & Chapman, T. 2004 Female resistance to male harm evolves in response to manipulation of sexual conflict. Evolution 58, 1028-1037.