Winner-loser effects on life history traits

Ageing of adult males could be accelerated by both high mating/reproductive effort and fighting for mates. Testing the relative importance of these factors is challenging, however, because males that win fights also tend to have more mates. We used a 2 x 2 experimental design to test how a prolonged (9 week) period of either winning or losing fights, and either high or low reproductive effort (manipulating by varying access to females) interact to affect male ageing and future reproduction allocation in the mosquitofish, Gambusia holbrooki. We measured telomere length and several life-history traits, including mating effort and ejaculates (sperm count and velocity). After 9 weeks there were significant differences between winners and losers in their mating effort, but not in their ejaculates. Males with a higher past reproductive effort (i.e. access to females) had significantly lower current mating effort and grew more slowly. Males with a higher past reproductive effort also had slower swimming sperm, but only if they were smaller than average in body size. Surprisingly, neither males with a higher past reproductive effort nor males that repeatedly lost fights had shorter telomeres. Our findings show that past social dynamics affect how males allocate resources to reproduction and somatic maintenance.

Methods
Creating winners and losers 

We randomly selected focal males to assign to experimental treatments and followed them individually throughout the study (n = 176). Focal males were anaesthetized briefly to measure their standard body length (SL) and then marked with a subcutaneous elastomer tag (NorthWest Marine Technology, WA, USA). After one week of isolation, focal males were randomly assigned to be winners or losers by being paired with either a smaller or larger rival. This method controls for intrinsic differences in fighting ability between males that might otherwise determine contest outcomes or affect investment into reproduction and somatic maintenance. Contests took place in 6 L aquaria that contained gravel, plastic plants for refugia, and had black plastic on three sides to minimize disturbance. Focal and rival males freely interacted for one week, after which a female was introduced to the tank to apply the reproductive treatment for another 8 weeks. In half the tanks, the two males could freely interact with the female and therefore the focal male could fully invest in reproduction (i.e. chase the female and mate). In the other half of the tanks there was also a female present, but she was kept behind a mesh barrier to prevent the males from mating with her. Equal numbers of winner and loser focal males therefore either had full access to a female (‘contests and mating’) or only interacted with a female through the barrier (‘contest only’) (NB: ‘contest’ refers to the fact that all males competed with a rival for the full 9 weeks). Rival males and stimulus females were rotated every 3-5 days so that focal males continued to fight to establish dominance, and to approach females

Male mating behaviour

After 9 weeks, each focal male was placed in a new 6 L aquaria with a random stock female. Each female was used once. Male mating behaviour was observed for 20 mins. We recorded: (a) time spent near the female (<5 cm and facing her); (b) the number of mating attempts; and (c) the number of successful attempts (i.e., those with the potential to transfer sperm).

Ejaculates

Immediately following the mating behaviour trials, focal males were anaesthetised in ice slurry to measure their body length (SL) and to strip their sperm. Males were then isolated for 5 days in 1 L aquaria to replenish their sperm reserves, after which we again stripped them. We had two ejaculate measures: total sperm count and sperm velocity.

For sperm counts, we vortexed the sample to disperse sperm then pipetted 3 µL onto a 20 µm capillary slide (Leja), and used a CEROS Sperm Tracker (Hamilton Thorne Research, Beverly, MA, USA) to count sperm under x100 magnification. We randomly counted five subsamples per sample. We estimated total sperm counts by adding the average sperm number per bundle to account for the six bundles removed for sperm velocity analyses.

To measure sperm velocity, we took two samples of three sperm bundles from each male’s ejaculate and pipetted the bundles into two separate PCR tubes containing 2 µL extender medium. We then pipetted each sample onto a cell of a 12-cell multi-test slide (MP Biomedicals, Aurora, OH, USA) coated with 1% polyvinyl alcohol solution (PVA). Sperm was ‘activated’ with 3 µL of solution (125 mM KCL and 2 mg/mL bovine serum albumin). We used a CEROS Sperm Tracker to record two measures of sperm velocity: VAP (average path velocity) and VCL (curvilinear velocity).

Relative telomere length

Relative telomere length (rTL) was measured using real-time quantitative PCR, determined as the ratio (T/S) of telomere repeat length (T) to a single-copy reference gene length (S). We used standard telomere primers Tel1b (5′-CGGTTTGTTTGGGTTTGGGTTTGGGTTTGGGTTTGGGTT-3′) and Tel2b (5′-GGCTTGCCTTACCCTTACCCTTACCCTTACCCTTACCCT-3′) and a Gambusia-specific region of the melanocortin 1 receptor (MC1R) as our control single-copy reference gene with the primers MC1R.F (5’-CCTGTAGGCGTAGATGAGCG-3’) and MC1R.R (5’-CACCAGTCCCTTCTGCAACT-3’). 

We ran qPCRs for each sample in triplicate on 96-well plates. Telomere and MC1R amplifications were run concurrently on separate plates using QuantStudio3 (Thermo Fisher Scientific, Waltham, USA). Each plate had three negative controls (9 µL reagent mix and 1 µL MilliQ purified water), two inter-plate control samples (run in triplicate, the same two individuals across all plates), and a golden sample at five DNA concentrations (0.05, 0.2, 1, 5 and 20 ng/µL) to generate the standard curve and determine the amplification efficiency of each plate (telomere: 1.99-2.11; MC1R: 1.93-2.01).