Nematode strains and culture conditions
Nematodes were cultured at 20 °C in Petri dishes containing NGM agar spotted with the Escherichia coli strain OP50 using standard protocols [50]. Geographic isolates of both C. elegans and C. remanei were raised under standardized laboratory generations for >10 generations to minimize maternal or laboratory adaptation effects during experiments. For normal maintenance, stocks were transferred twice weekly to prevent overcrowding and dauer formation. Strains were frozen at −80 °C in a glycerol solution using standard procedures [50].
In C. remanei, phenotypes were compared among three geographic isolates: PB259, EM464, and SB146. Both SB146 and EM464 were obtained from the Caenorhabditis Genetics Center (CGC), whereas PB259 was a gift from Scott Baird at Wright State University. These are largely isofemale lines and may have been subject to some inbreeding during derivation, although they are known to harbor substantial residual genetic variation.
In C. elegans, twelve geographic isolates served as the initial pool of standing genetic variation for experimental evolution in the laboratory. Strain names, along with the location where originally isolated, were as follows: AB1 and AB2 from Australia; PB303, PB305, PB306, and PB307 from Ohio; CB4855, CB4857, and DH424 from California; CB4856 from Hawaii; CB4932 and N2 from England. Most of these strains were obtained from the CGC, whereas the four “PB” strains were gifts from Scott Baird at Wright State University. Two additional CGC strains were also used: JK574 served as the source of the fog-2 (q71) mutation that renders hermaphrodites self-sterile [30]; this mutation was used to prevent self-fertilization in half of the evolving lines (see below). PD4792 served as the source of the transgenic array that expresses high levels of Green Fluorescent Protein (GFP) in the pharynx; this transgene was used to determine paternity in sperm competition experiments involving C. elegans.
Experimental evolution of C. elegans
The genetic variation present across the 12 geographical isolates was mixed together to serve as the founding populations for the creation of both High-Competition (HC) and No-Competition (NC) lines for experimental evolution. To create the founding stock for the HC lines, each geographic isolate was first modified to render hermaphrodites self-sterile (i.e., unable to produce their own sperm, thereby needing males in order to reproduce). This was accomplished by introgressing the fog-2 (q71) mutation from stock JK574 into each genetic background independently via ten generations of backcrossing. The fog-2 (q71) mutation was maintained during these introgressions by selecting for self-sterile hermaphrodites every few generations. Worms from the resulting [fog-2] introgression strains were then mixed in equal proportions and maintained in a large population to mate freely for at least six generations, in order to recombine the genomes and thereby create the HC-founding stock. To create the NC-founding stock in parallel, self-fertile hermaphrodites and males from the original twelve geographic isolates were mixed in equal proportions and maintained in a large population to mate freely for at least six generations, in order to recombine the genomes and thereby create the NC founding stock. Both the HC-and NC-founding stocks were frozen for later analysis.
For experimental evolution of HC and NC lines, we essentially replicated the protocol used by [41]. For each of three independent HC lines, 60 L4 self-sterile hermaphrodites and 100 L4 males were picked to found each generation, beginning with worms from the HC-founding stock. In parallel, for each of three independent NC lines, 60 L4 self-fertile hermaphrodites were picked to found each generation, beginning with worms from the NC-founding stock. Both HC and NC lines were allowed to evolve in this manner for 60 generations, with worms frozen for later analysis at generations 30 and 60. After experimental evolution was halted, worms from the evolved lines were thawed and examined for the following male phenotypes: sperm size, sperm competitive ability, male mating behavior, and effects of male exposure on hermaphrodite mortality rates.
Life history assays
To determine population-specific effects of mating on female lifespan in C. remanei, the survivorship of age-synchronized cohorts of virgin and mated females was compared, with mated females receiving either a short-term (24 h) access to males, or with females having life-long access to males. For unmated/virgin longevity, egg-laying females from each strain were picked randomly from the stock plates to fresh 90-mm Petri dishes (30 females per plate, three plates per strain) for three hours to obtain age-synchronized offspring, and then removed. When the synchronous offspring had molted to L4 larvae, virgin females were identified and transferred to individual 30-mm dishes (65 EM464 females, 40 PB259 females, and 60 SB146 females, respectively). Individuals were transferred daily to fresh dishes and checked for mortality. A female was scored as dead when it showed no signs of movement, no pharyngeal pumping (a feeding behavior), and did not respond to light prodding with a fine platinum wire.
For the short term (24 h) mating assays, procedures were as above except that when individuals had molted to the L4 stage, they were identified to sex, and 5 males and one female were each placed on a 30-mm Petri dish. All possible combinations of the three strains were crossed, to examine possible interaction effects between males and females of different genetic backgrounds. All males were removed after 24-h, and the females then transferred daily during their egg-laying period, and less frequently in their post-reproductive period [37]. Daily mortality was scored as above. For the long term mating assays, males were maintained throughout the female lifetime. Extra males from the age-synchronized plates were supplied as needed to replace dead or injured males to maintain constant male density, and plates were inspected daily for female mortality.
For all of the crosses described above, female fecundity was calculated by reserving all plates containing eggs and counting the offspring under a dissecting microscope. To facilitate this and diminish counting errors, these plates were incubated at 20 °C until the offspring were at least L4 larvae in size. Because of this delay in counting, any inviable eggs or juvenile mortality were discounted in estimates of female fecundity. Mating effects were estimated using JMP 10 (SAS Institute) via a 3-way factorial ANOVA with female strain, male strain, and length of male access as fixed effects. Posthoc tests were conducted using Tukey’s Honest Significant Difference (HSD) procedure.
Male effects on hermaphrodite longevity within C. elegans were assessed using a mass-mating assay in which ten wildtype hermaphrodites were placed with 20 males from one of the following sources: an HC line, an NC line, CB4856 (Hawaiian isolate), N2 (Bristol, England isolate), the original mixed stock from which the HC lines were founded, or the original mixed stock from which the NC lines were founded. The investigator was always blind to the source of the males. Every 24 h for 5 days, all living worms were transferred to fresh plates, any dead or missing males were replenished from source plates, and daily hermaphrodite deaths were scored. Differences in longevity were assessed using an ANOVA, with comparisons to the ancestor made via Dunnett’s procedure.
Sperm size
For at least two generations prior to sampling sperm, worms were kept on uncrowded plates to insure abundant food and avoid dauer formation. Virgin L4 males were isolated 20–24 h prior to dissection, and dissected according to a standardized protocol [51] with electrolytically sharpened tungsten needles (C. elegans) or insect pins (C. remanei) in freshly thawed aliquots of SM1 adjusted to pH 7.0 and supplemented with 10 mg/ml polyvinylpyrrolidone (Sigma PVP40). Spermatid images were captured using Nomarski optics at 600X (C. elegans) or 400X (C. remanei) magnification. Cross-sectional areas of round spermatids (i.e., chosen prior to any shape changes associated with spermatid activation) were measured using NIH Image or Image Pro (Media Cybernetics). For C. remanei, an average of 60 spermatids from ten different males were measured from each of the three different strains to estimate among-strain differences. For the C. elegans experimental evolution lines, ten spermatids were measured from each of 20 males (i.e., 200 spermatids total) at each of three different times points (generations 0, 30, and 60). Differences in sperm size were analyzed via an analysis of variance (ANOVA) using JMP 10 (SAS Institute). For C. remanei, strain was the sole main effect. For C. elegans, a nested ANOVA model including competition treatment, replicate line (nested within competition), and generation was fit, with competition and generation as fixed effects, and replicate line as a random effect. Specific evolutionary hypotheses were tested using contrast coefficients as a subset of the overall model.
Sperm competition
In C. elegans, self-sterile hermaphrodites from stock CB4856 harboring [fog-2] introgessions were used as the “arena” for male-male competition. To create standard competitor males, the [mIsll IV] transgene from stock PD4792 was introgressed into the AB2 stock genetic background by 10 generations of backcrossing with selection to maintain the array; worms harboring the transgene express Green Fluorescent Protein (GFP) in the pharynx muscles, which served as a dominant marker to score paternity in progeny (i.e., worms expressing GFP were sired by AB2 [GFP] males, whereas worms lacking GFP expression were sired by the males being tested for competitive ability).
During the two generations prior to each sperm competition assay, all of the relevant stocks were refreshed on new plates at low densities to make sure that worms were consistently well-fed and did not spend any time as dauer larvae. Three days before the sperm competition assay began, 30–50 eggs from each stock were transferred to empty, fresh plates so that they could develop at controlled densities. About 24 h prior to the beginning of the assay, worms at a standardized L4 stage were moved to fresh plates and kept at low densities, away from members of the opposite sex, in order to control mating opportunities prior to the sperm competition assay.
To test sperm offense ability, one CB4856 [fog-2] self-sterile hermaphrodite was exposed to 5 AB2 [GFP] males for 24 h, then transferred to a new plate with 5 HC or 5 NC males for another 24-h (Day 1 plate). The hermaphrodite was then kept on a plate alone for 24-h (Day 2 plate) and then moved again to a fresh plate for a final 24-h period (Day 3 plate). Any hermaphrodite dying before the end of the experiment was excluded from the data set. Once the progeny on the Day 1 – Day 3 plates had reached maturity, the ratio of GFP to non-GFP progeny was counted (a minimum of 30 random progeny were counted for each plate per day) with the investigator blind to the source of male parents. For sperm defense, the same experimental design was employed, only switching the HC or NC males to the first day and the AB2 [GFP] males to the second day.
In C. remanei, sperm competition assays were carried out essentially the same as above, except that C. remanei females were used as the “arena” for male-male competition, and males from the three strains were allowed to compete directly (i.e., no standardized male competitors), with offspring paternity determined by examining a microsatellite marker that varied among strains. Specifically, a tetranucleotide microsatellite, (GACA)13, was identified from contig 52.84, indices 6754–6807, of the C. remanei Pcap Asembly v2 (Genome Sequencing Center, Washington University School of Medicine, St. Louis, Missouri, USA) using Tandem Repeats Finder v 3.21 [52]. Primers flanking this repeat, GGAACAGATGAGGTGATGACG and CATCTCCGCTCTCCAATGA, were designed in Primer Designer v 2.0 (Scientific and Educational Software, Cary, NC). EM464 and SB146 are fixed for different alleles at microsatellite locus cr52.84, approximately 210 bp and 180 bp respectively, allowing paternity assignment of F1 progeny resulting from crosses between these strains.
Sperm competitive ability was measured both as the ability of a focal male to displace the sperm of the first male when introduced as the second male (“offense” or P2; [53]) and to avoid displacement of sperm from the second male when introduced as the first male (“defense” or P1). These were not measures of single mating sperm displacement, since it was likely that the males and females mated multiple times during the day that they were together. “Day 1” of sperm competition was the first day when both types of competing sperm were present within the same female. On days 2–3, no males were present on the plates with females, so any offspring produced on these days were generated from residual sperm remaining from the initial matings. For C. remanei, male-specific effects were tested by comparing P1 and P2 measures in a particular cross combination in order to estimate whether males from a particular strain performed better as first and/or second males compared to those of another strain. This effect was tested using Fisher’s Exact Test of the sperm counts for each male. For C. elegans, sperm competitive ability relative to the fixed GFP-tester line was compared between the high-competition and no-competition lines using offspring counts within a nested logistic-regression that tests the effect of the experimental treatment while taking into account variance among experimental replicates (JMP 10, SAS Institute).
C. elegans male mating behavior
One virgin male from either an HC or NC line was placed on an OP50 bacterial spot (~7 mm diameter, grown overnight) with ten virgin, four-day-old wildtype hermaphrodites from N2 Bristol, and observed for twenty minutes. The observer was blind to the source of the male. Successful matings within twenty minutes were noted, and for these matings the time until spicule insertion and the duration of spicule insertion were recorded.
Availability of supporting data
The data sets supporting the results of this article are available in the Dryad Data Repository (www.datadryad.org), doi:10.5061/dryad.b13f4.