PreprintPDF Available

Meiotic drive does not impede success in sperm competition in the stalk-eyed fly, Teleopsis dalmanni

September 2022

DOI:10.1101/2022.09.09.507334

License
CC BY-NC-ND 4.0

Authors:

Sadé Bates

University College London

Andrew Pomiankowski

University College London

In male X-linked meiotic drive systems, the driver causes degeneration of Y-bearing sperm, leading to female-biased offspring sex ratios. This potentially leads to a two-fold transmission advantage to drive chromosomes. However, drive-bearing sperm often do poorly in sperm competition, limiting their ability to spread. We use the stalk-eyed fly, Teleopsis dalmanni , to investigate the success of the X-linked Sex Ratio (SR) meiotic drive system. In this species, polyandrous matings, where a female mates with multiple males, are common. Recent findings demonstrate SR males transfer the same numbers of viable sperm as wildtype (ST) males during mating, implying that they do not necessarily have reduced fertility under sperm competition. Reciprocal mating trials were performed to measure the success of SR and ST sperm in double mated females, with either a SR or ST male mated first followed by a male of the alternative genotype. There was no significant difference in the number of offspring sired by SR and ST males. This equivalence held regardless of whether the SR male mated first or second. We show these results are consistent with previous studies that suggested SR male sperm do poorly in sperm competition. Future experiments will determine whether the competitive ability of SR males is maintained under higher stress conditions likely to be experienced in nature, in which females mate repeatedly with multiple males. The results from the current study helps to explain the high meiotic drive frequency of around 20% in wild populations in this species. Impact Summary Meiotic drive genes are selfish genetic elements that distort Mendelian patterns of inheritance to bias transmission in their favour. We use the stalk-eyed fly, Teleopsis dalmanni , to investigate the fitness effects associated with a meiotic drive gene called Sex Ratio (SR), which is linked to the X chromosome. In males, SR destroys Y-bearing sperm, meaning only X-bearing sperm are viable, and females who mate with drive males sire all-female broods. This confers a two-fold transmission advantage to the SR gene, as it is transmitted to all offspring. We recently discovered that drive males have evolved compensatory mechanisms to cope with the sperm destruction caused by meiotic drive. They have greatly enlarged testes, allowing them to produce more sperm. When drive males mate with females, they deliver as many sperm and sire as many offspring as wildtype males. Building on this finding, we measured how drive male sperm performs against sperm from a non-carrier male in sperm competition – where the sperm from different males compete to fertilise an egg. Double mating trials were performed, where a single female was mated once to a drive and once to a non-carrier male. By genotyping offspring, we show that the number of offspring sired by the drive male was not different from the number sired by the non-carrier competitor. These findings contrast with those in other species. Typically, drive males do poorly in sperm competition and their spread is severely restricted by sperm competition. In stalk-eyed flies, female multiple mating with many males is the norm, but this does not appear to inhibit the fertility of drive males. The success of drive under sperm competition helps to explain the high frequency of drive around 20% in natural populations of T. dalmanni .

Content uploaded by Andrew Pomiankowski

Content may be subject to copyright.

Available via license: CC BY-NC-ND 4.0

Content may be subject to copyright.

Evolution, 2023, 77(10), 2326–2333

https://doi.org/10.1093/evolut/qpad149

Advance access publication 24 August 2023

Brief Communication

Meiotic drive does not impede success in sperm

competition in the stalk-eyed ﬂy, Teleopsis dalmanni

SadéBates1,, LaraMeade1,, AndrewPomiankowski1,2,

1Department of Genetics, Evolution and Environment, University College London, London, United Kingdom

2CoMPLEX, University College London, London, United Kingdom

Corresponding author: Department of Genetics, Evolution and Environment, University College London, Gower Street, London, WC1E 6BT, United Kingdom.

Email: ucbhpom@ucl.ac.uk

Abstract

Male X-linked meiotic drive systems, which cause the degeneration of Y-bearing sperm, are common in the Diptera. Sperm killing is typically

associated with ﬁtness costs that arise from the destruction of wildtype sperm and collateral damage to maturing drive sperm, resulting in poor

success under sperm competition. We investigate X-linked meiotic drive fertility in the stalk-eyed ﬂy, Teleopsis dalmanni. Drive male paternity

was measured in double mating trials under sperm competition against a wildtype male. Drive males sired the same number of offspring as

wildtype males, both when mated ﬁrst or second. This is the ﬁrst evidence that drive males can compete equally with non-drive males in double

matings, challenging the assumption that drive males inevitably suffer reduced fertility. The ﬁnding is in accord with previous work showing that

the number of sperm per ejaculate transferred to females during non-competitive single matings does not differ between drive and wildtype

males, which is likely due to the adaptive evolution of enlarged testes in drive males. Future experiments will determine whether the competitive

ability of drive males is maintained under higher rates of female remating likely to be experienced in nature.

Keywords: meiotic drive, stalk-eyed ﬂy, sperm competition, multiple mating

Introduction

Meiotic drive causes the unequal transmission of genes to

the next generation, violating Mendelian laws of segrega-

tion (Gershenson, 1928; Sandler & Novitski, 1957). In the

extreme, the driver entirely excludes wildtype alleles and is

transmitted to all offspring (Searle & de Villena, 2022; Wolf

et al., 2022). X-linked drivers are common among Diptera

species and lead to dysfunction of Y-bearing sperm and the

production of female-only broods (Hurst & Pomiankowski,

1991; James & Jaenike, 1990; Jiggins et al., 1999; Newton et

al., 1976; Policansky, 1974; Presgraves et al., 1997). Such a sig-

nicant transmission advantage could potentially lead to pop-

ulation extinction due to the lack of males (Hamilton, 1967;

Hatcher et al., 1999; Mackintosh et al., 2021). However, the

tness costs associated with carrying drive genes often result

in negative frequency-dependent selection, which limits their

spread (Finnegan et al., 2019; Lindholm et al., 2016).

One factor that strongly impacts the spread of meiotic drive

genes is reduced fertility (Zanders & Unckless, 2019). Males

with drive not only lose wildtype gametes but typically suf-

fer pleiotropic “collateral damage” that reduces the activity

or number of mature drive sperm, leading to poor outcomes,

especially under sperm competition (Price & Wedell, 2008).

This decit is likely to be prominent in insects that possess

reproductive organs specialized for long-term storage of via-

ble sperm, increasing interactions between ejaculates (Parker,

1970). Evidence from sperm competition studies of X-linked

meiotic drive systems in Drosophila species supports this pre-

diction. In Drosophila pseudoobscura, SR drive males sire

fewer offspring than standard males in double mating trials

(Price et al., 2008a). Drive males have a disproportionally

lower success both in their ability to defend against other

sperm as the rst (P1) male or to displace sperm already in

storage as the second (P2) male (Price et al., 2008a). A similar

pattern occurs in Drosophila simulans with reduced success

in P1 and P2 positions for drive males, and preferential drive

male sperm ejection from the female reproductive tract even

without competition from the second male’s sperm (Angelard

et al., 2008; Atlan et al., 2004). It has been suggested that in-

creased female polyandry evolves to undermine the success of

drive sperm and an experimental evolution study in D. pseu-

doobscura and a double mating experiment in Drosophila

recens support this possibility, linking the frequency of drive

with the rate of multiple mating (Courret et al., 2019; Dyer

& Hall, 2019; Haig & Bergstrom, 1995; Price et al., 2008b;

Zeh & Zeh, 1997).

In this article, we investigate the association between

X-linked meiotic drive and reduced male fertility using

the X-linked SR meiotic drive system in the stalk-eyed y,

Teleopsis dalmanni. Stalk-eyed y females store sperm in

the spermathecae (long-term storage organs) after mating,

before sperm migrate to the ventral receptacle, where they

are individually packaged into pouches prior to release into

the oviduct for fertilization of mature eggs (Kotrba, 1995;

Presgraves et al., 1999). In several stalk-eyed y species, the

main mode of sperm competition is sperm mixing, rather than

male precedence (Bellamy, 2012; Corley et al., 2006; Lorch

et al., 1993; Wilkinson et al., 1998a). Double mating trials

Received October 10, 2022; revisions received July 14, 2023; accepted August 21, 2023

Associate Editor: Brandon S.Cooper; Handling Editor: TraceyChapman

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/),

which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2327Evolution (2023), Vol. 77, No. 10

appear to conrm that drive males should be poor competi-

tors as drive (SR) males sired fewer offspring than wildtype

(ST) males (Wilkinson & Fry, 2001; Wilkinson et al., 2006).

However, several factors raise concerns about a simplistic

interpretation of these ndings. The rst study was an in-

ter-population cross of Malaysian and Thai ies. It was car-

ried out before genetic markers had been developed and used

variation in leg color to assign parentage, which has an un-

known error rate (Wilkinson & Fry, 2001). In addition, this

study was in the conger Teleopsis whitei which may well have

a different pattern of sperm competition than in T. dalmanni.

The second study is in T. dalmanni and reported a lower SR

male paternity using double mating trials (Wilkinson et al.,

2006). However, this effect was limited to broods in which

all offspring were sired by a single parent, that were less fre-

quently fathered by SR males. There was no difference in SR

and ST paternity in mixed paternity broods. In addition, this

experiment only considered the competitive ability of SR

males when mating second. This means that defensive traits

of SR sperm and ejaculate were not assessed, so it is unclear

whether the lack of success of SR males is general or limited

to lower sperm precedence when mating second.

In addition, a profound challenge arises from recent nd-

ings that SR males transfer similar numbers of sperm per

ejaculate (Meade et al., 2019, 2020). This was measured in

females both in the spermathecae and the ventral receptacle

after matings with SR or ST males, as well as after up to three

sequential matings by a single male (Meade et al., 2019).

Furthermore, when egg counts were used to measure fertil-

ity after single matings, it did not differ for females mated

to SR or ST males (Meade et al., 2020). Dissection of adult

SR males reveals that they have greatly enlarged testes, which

allow sperm delivery and fertility to be maintained despite

the destruction of sperm caused by meiotic drive (Meade et

al., 2019, 2020). This challenges the conventional view that

drive males are weak competitors, and specically, the nding

of a competitive decit of drive males in double mating trials.

Here, the competitive success of SR males was measured in

a standard sperm competition assay using reciprocal double

mating trials in which the SR male mated rst followed by

the ST male, or vice versa. This allowed an assessment of the

SR male’s success in both the offensive and defensive role and

revealed whether there is rst or last male sperm precedence.

Even though multiple mating well above two is the norm in

T. dalmanni stalk-eyed ies (Baker, 2001; Baker et al., 2003;

Chapman et al., 2005), the simplicity of the double mating

trial allows clear assessment of whether SR sperm suffer a

disadvantage in competition with ST sperm when the two

males mate equally. The offspring arising from these trials

were collected and genotyped at the larval stage to determine

the proportion of offspring sired by SR males. This enabled

the study to avoid confounds in paternity share relating to egg

to adult viability differences, which have recently been shown

to disadvantage SR-carrying larvae (Finnegan et al., 2019).

Methods

Stock populations

Flies for the standard stock (ST-stock) population carry

only the wildtype X chromosome (XST). They were collect-

ed (by S. Cotton and A. Pomiankowski) in 2005 from the

Ulu Gombak valley, Peninsular Malaysia (3°19ʹN 101°45ʹE).

They have since been maintained in high-density cages (>200

individuals) to minimize inbreeding and are regularly moni-

tored to ensure they do not contain the meiotic drive.

The meiotic drive stock (SR-stock) population is composed

of females that are homozygous for a sex-ratio-distorting X

chromosome (XSR). They were derived from ies collected in

2012 (by A. Cotton and S. Cotton) from the same location

as the ST-stock. XSR/Y males produce 100% female offspring

due to transmission distortion. The XSR female stock is main-

tained by crossing XSR/XSR females with XST/Y males to pro-

duce XSR/Y drive males, who are then mated to the XSR/XSR

females to generate the next generation of the SR-stock fe-

males. The outcrossing to ST males from the ST-stock ensures

that the two stocks only differ in their X chromosomes and

are homogenized for autosomal content.

Both stock populations were kept at 25 °C, with a 12:12

hr dark:light cycle and fed puréed sweetcorn twice weekly.

Fifteen-minute articial dawn and dusk periods were created

by illumination from a single 60W bulb at the start and end

of the light phase.

Experimental populations

Experimental ST (XST/Y) and SR males (XSR/Y) were drawn

from the ST-stock and SR-stock, respectively. They were

housed separately in cages of ~50 individuals until sexually

mature, in groups of similar age (6–8 weeks). ST-stock females

were added to these cages at an equal sex ratio for > 3 days

to allow males to mate. The females were then removed and

discarded. Experimental males were then kept in single-sex

groups for a further 3–6 days to allow their accessory glands

to return to full size (Rogers et al., 2005).

Experimental ST females (XST/XST) were drawn from the

ST-stock. All experimental females were virgins, 6–8 weeks

old, and had reached sexual maturity (Baker et al., 2003).

ST females were anesthetized on ice and their eyespans were

measured (see below method) to exclude small ies and limit

variation in size and fecundity that could inuence sperm al-

location strategies in males (Cotton et al., 2015). Only large

females with an eyespan > 5.4 mm were used in mating trials

(range 5.4–5.8 mm).

Sperm competitiveness of SR and ST males

Mating trials were conducted to measure the competitiveness

of SR and ST males. On the day preceding each assay, ex-

perimental females were housed singly in 500 ml clear plas-

tic containers with a moist cotton wool base. On the trial

day, a single male was added to each container ~15 min after

dawn, as this is the period during which mating is most likely

(Chapman et al., 2005). Males were allowed to mate, dened

as a copulation lasting ≥ 30 s, as durations shorter than this

are usually insufcient for sperm transfer (Cotton et al., 2015;

Rogers et al., 2006). The mating duration was recorded. If no

mating was observed after 15 min, the male was moved to a

new container with a new female. If mating still did not occur

after a further 15 min, the male was discarded. The origi-

nal unmated female was used again and placed with another

male. If this did not result in a copulation after 15 min, the

female was discarded.

A second mating was performed 24 hr later, following the

same protocol. Again, if the male failed to copulate with the

female after 30 min, he was replaced, and if a mating still did

not occur, the female was discarded. The mating failure rate

was extremely low: one ST male failed to mate on day 1 (P1),

one SR male failed to mate on day 2 (P2), and one female was

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2328 Bates et al.

discarded as she failed to mate with any male. Females were

mated either to an SR male followed by an ST male, or an ST

male followed by an SR male. Once females had been double

mated, the containers were lined with a fresh moist cotton

wool base and 1tsp puréed sweetcorn to collect eggs, which

was renewed every 2–3 days for 2 weeks. This kept larval

density low, maximizing survival. Bases were stored in Petri

dishes at 25 °C. In total, 62 females were successfully mated

twice: 30 to an SR male rst and 32 to an ST male rst. For

ease, these matings were carried out in two batches, 1 week

apart.

After mating, experimental males were removed and fro-

zen, and their eyespan and thorax length were measured

under a Leica microscope using ImageJ (v1.46; Schneider

et al., 2012). Eyespan was dened as the distance between

the outer tips of the eyes (Hingle et al., 2001). Thorax length

was dened as the distance ventrally from the anterior tip

of the prothorax along the midline, to the joint between the

metathoracic legs and the thorax (Rogers et al., 2008).

Progeny genotyping

Petri dishes were examined for larvae one week after collec-

tion. Larvae that had developed to be large enough to be seen

by eye were transferred to a 96-well plate. Each Petri dish

was then examined daily to collect the remaining growing

larvae until there was no further evidence of their presence.

Each well of the plate contained 100 µl digestion solution (20

mM EDTA, 120 mM NaCl, 50 mM Tris-HCL, 1% SDS, pH

8.0) and 4 µl proteinase K (10 mg ml−1). A standard proto-

col was adapted to extract and purify DNA from larvae (see

Supplementary Information 1 for details; protocol from Burke

et al., 1998). The X-linked INDEL marker comp162710 was

used to identify offspring of ST and SR fathers, due to its

reported accuracy in determining phenotype (>90%; Meade

et al., 2019). XST carries a large allele (286 bp), whereas XSR

carries a small allele (201 bp).

Nine females produced no offspring. A further two females

produced low numbers of offspring (2, 6), of which none and

one were successfully genotyped, respectively. Overall, in 7

of 31 cases, the mating order was P1 ST—P2 SR, and in 4 of

31 cases, the mating order was P1 SR—P2 ST. There was no

mating order effect on failure to produce genotyped offspring

(Fisher exact test p = .508). These 11 females were removed

from further analysis.

Not all offspring collected over the 2-week period were

genotyped for logistical reasons. On average, 39.8 (range

0–116) offspring were collected, and 21.9 (range 0–59) off-

spring were genotyped per female; a total of 1,161 successful

PCRs. The 96-well plates were genotyped without regard to

the offspring of particular females as they were collected on

particular days. This approach led to a high correlation be-

tween offspring production and genotyping (ρ = 0.872, n =

51, p < .001).

Statistical methods

All tests were carried out in R version 4.1.2 (R Core Team,

2021). To test if mating order or genotype affected the number

of offspring sired by each male, P1:P2 offspring (the number of

offspring sired by P1 relative to the number of offspring sired

by the P2 male) or ST:SR offspring (the number of offspring

sired by the ST relative to the number of offspring sired by the

SR male) were tted as the response variable in generalized

linear models (GLMs) with a binomial error distribution. The

response variables were coded using the R cbind function.

Count data of offspring sired by each male was used in the

binomial analysis rather than one male’s paternity proportion

to account for the variable sample size of offspring assigned

to each male (larger sample sizes provide better estimates), as

used by others (Dobler et al., 2022). It is not possible to treat

mating order and genotype in a single “global” model com-

bining genotype and mating order as each female’s offspring

are derived from only two males who have both a genotype

and mating order. Hence, the binomial analysis (y1, y2) enters

offspring either according to mating order (y1 = P1, y2 = P2)

or genotype (y1 = ST, y2 = SR) in two separate analyses. As

the GLMs were over-dispersed, a quasi-binomial error distri-

bution was used. Tests were repeated excluding females that

had ≤10 offspring genotyped. The number of larvae collect-

ed and the batch in which the matings were performed were

assessed as potential confounding variables. In addition, the

data were split in two, considering offspring number of SR/

ST or in the P1/P2 role, with linear models on genotype. In

order to assess the power of the experiment to detect differ-

ences in mating order or genotype, the same GLM statistic

was calculated with up to a 10-fold increase in sample size

on re-sampled data (with replacement). One thousand repeats

were performed at each sample size, and the resulting GLM

statistics examined for evidence of difference in paternity due

to mating order or genotype (see Supplementary Information

4 for detailed method description and code).

The effect of male thorax length (a proxy for body size) and

relative eyespan (the variation in eyespan after controlling

for thorax length) were also considered in the analysis. Both

traits are strongly condition dependent and indicators of male

genetic and phenotypic quality (Cotton et al., 2015; David et

al., 2000; Howie et al., 2019). Whether these male trait sizes

differed between genotypes was tested by tting thorax length

and relative eyespan as the response variable in linear mod-

els. In addition, whether mating duration differed by mating

order and genotype was tested by tting mating duration as

a response variable in linear models, and by its inclusion as

a xed effect in GLMs with the number of offspring sired

by each male. Full statistical analyses are reported in the

Supplementary Informations 2 and 3.

Results

Male fertility

In total, 62 females were reciprocally mated to males of each

genotype. Fifty-one females had offspring (between 4–59)

that were successfully genotyped (27 P1 SR—P2 ST and 24

P1 ST—P2 SR matings), and of these, 47 females had ≥10

genotyped offspring (23 P1 SR—P2 ST and 24 P1 ST—P2

SR). For two of the reciprocal matings, one mating was 29

s in duration; these matings were included in the subsequent

analysis as, in both cases, the male in question produced off-

spring.

The distribution of proportions sired by the two males

was at, including offspring broods that were exclusive-

ly sired by either the P1 or P2 male (Figure 1A) or by ei-

ther the ST or SR male (Figure 1B), with means around

equality (mean ± SD P2 male = 0.522 ± 0.327, SR male

= 0.575 ± 0.316). Using offspring numbers (rather than

proportions), there was no effect of mating order (F1,49 =

1.307, p = .259; Figure 2A) or genotype (F1,49 = 0.196, p =

.660; Figure 2B) on the number of offspring sired by each

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2329Evolution (2023), Vol. 77, No. 10

male. Nor was there an effect of genotype when the data

were split in halves, either with the SR male in the P1 role

(F1,49 = 0.002, p = .963), or in the P2 role (F1,49 = 0.434,

p = .513). An additional test added the total number of

offspring collected as a covariate as it varied between fe-

males (mean ± SD; 48.196 ± 22.735 offspring; range 6–116

offspring), but its inclusion did not alter the main effects

of mating order or genotype (p > .05; see Supplementary

Information 2, Figure S2). Likewise, the main effects were

unchanged when batch number was included as a covariate

(p > .05; see Supplementary Information 2). The results of

these tests were also unchanged after the exclusion of the

four females that had less than 10 offspring genotyped (see

Supplementary Information 3).

In 11 of the 47 cases with ≥10 offspring genotyped, one

male sired more than 0.95 of the offspring, with no difference

between male mating position (four sired by the P1 male, and

seven sired by the P2 male, F1,9 = 0.986, p = .351) or male

genotype (eight sired by the SR male and three were sired by

the ST male, F1,9 = 0.841, p = .383). When these extreme cases

were excluded, there was still no effect of mating order (F1,32

= 0.094, p = .761) or male genotype (F1,32 = 0.589, p = .448)

on the number of offspring sired.

To assess the power of the data to detect differences, the

data were resampled (with replacement) using a 1–10-fold

increase in sample size compared to the original data (1,000

repeats for each fold increase, Supplementary Information 4).

As expected, the fraction of runs with signicant differences

Figure 1. (A) The distribution of P2, the proportion of offspring sired by the second male, is shown per brood (blue). (B) The distribution of the proportion

of offspring sired by the SR male is shown per brood (red).

Figure 2. (A) Points correspond to the number of P2 offspring against the total number of offspring per brood. The solid blue line represents the

regression of the number of P2 offspring against the total number of offspring (β = 0.539; intercept constrained to zero). The blue dashed line

represents P2 = 1.000 (all P2 offspring), the black dashed line represents P2 = 0.500 (equal P1 and P2 offspring), and the blue dotted line represents

P2 = 0.000 (all P1 offspring). (B) Points correspond to the number of SR offspring against the total number of offspring per brood. The solid red line

represents the regression of the number of SR offspring against the total number of offspring (β = 0.472; intercept constrained to zero). The red

dashed line represents SR = 1.000 (all SR offspring), the black dashed line represents SR = 0.500 (equal SR and ST offspring), and the red dotted line

represents SR = 0.000 (all ST offspring).

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2330 Bates et al.

(at p < .05) increased with sample size. The increase was

marked for mating order with a P2 advantage evident at a

4-fold increase in sample size (95% condence interval [CI]

0.207–4.567 in favor of P2). However, the increase was mi-

nor for genotype, and there was no advantage to either gen-

otype even with a 10-fold increase in sample size (95% CI

−0.789–3.667 in favor of SR).

Male trait size and mating duration

Thorax length was smaller in SR than ST males (mean ± SE,

SR = 2.190 ± 0.023 mm, N = 49, ST = 2.297 ± 0.025 mm, N =

50; F1,97 = 9.783, p = .002). Eyespan is strongly colinear with

thorax (F1,97 = 167.242, p < .001) and was likewise smaller in

SR males (SR = 7.304 ± 0.111 mm, ST = 7.897 ± 0.115 mm;

F1,97 =13.766, p < .001; Supplementary Information 2, Figure

S1). However, the relative eyespan did not differ between gen-

otypes (F1,96 = 3.734, P = 0.056; Supplementary Information

2, Figure S1). As thorax length differed between genotypes,

it was added as a covariate, but there was still no effect of

mating order (F1,44 = 1.161, p = .287) or male genotype (F1,44

= 0.369, p = .547) on the number of offspring sired by each

male.

Mating duration did not differ with mating order (mean ±

SE, P1 = 63.94 ± 3.43 s, P2 = 73.45 ± 3.43 s; F1,100 = 0.943,

p = .334) or genotype (ST = 60.82 ± 2.36 s, SR = 76.57 s

± 9.42 s, F1,100 = 2.627, p = .108). Mating duration did not

affect the number of offspring sired by the P2 male (F1,48 =

0.022, p = .882), but P1 males with a shorter mating duration

sired a greater number of offspring (F1,48 = 4.082, p = .049).

Mating duration did not affect the number of offspring sired

by the SR male (F1,48 = 0.246, p = .622) or the ST male (F1,48 =

3.366, p = .073). Given its inconsistent effect on the number

of offspring sired, the mating durations of the two males were

added as covariates, but there was still no effect of mating

order (F1,47 = 1.208, p = .277) or genotype (F1,47 = 0.071, p =

.791) on the number of offspring sired.

Discussion

Our study provides little support for the idea that males

carrying X-linked meiotic drive are at a disadvantage un-

der sperm competition due to sperm loss and other delete-

rious effects of meiotic drive on sperm function (Courret et

al., 2019; Verspoor et al., 2020). Here, the paternity of SR

males did not differ from ST males overall, nor in the P1 or P2

positions considered separately. This challenges the general

pattern which has been reported across the Diptera (Dyer &

Hall, 2019; Hurst & Pomiankowski, 1991; James & Jaenike,

1990; Jiggins et al., 1999; Newton et al., 1976; Policansky,

1974; Presgraves et al., 1997; Price et al., 2008a). It is also

in opposition to previous evidence of lower drive male pa-

ternity in stalk-eyed y double-mating experiments, which

were discussed in the Introduction (Wilkinson & Fry, 2001;

Wilkinson et al., 2006). Our results are robust to a number

of potential confounding factors: matings were performed be-

tween ies from the same population, offspring paternity was

assessed using highly accurate genetic markers, larvae were

used to assess paternity—which reduces the impact of lower

egg-adult viability in SR females—and double matings were

carried out with SR males in the rst and second mating po-

sition to reliably assess sperm precedence. Furthermore, the

ndings here align with those of Meade et al. (2019, 2020),

who showed that sperm numbers transferred to females and

the resulting fertility do not differ in single matings by SR and

ST males.

Our results do not invalidate previous ndings, which

likely reect genuine experimental differences. The study of

Wilkinson and Fry (2001) was carried out on the closely re-

lated species T. whitei, which also carries X-linked SR meiotic

drive that is thought to have evolved prior to the divergence

of these two species (Meier & Baker, 2002; Presgraves et al.,

1997). Genetic markers for drive have not been identied in

T. whitei (G. S. Wilkinson, personal communication), imply-

ing a small inversion is associated with drive in this species,

unlike the multiple inversions that cover most of the T. dal-

manni SR X chromosome (Christianson et al., 2011; Paczolt

et al., 2017; Reinhardt et al., 2014, 2023; Wilkinson et al.,

2006). This means that few X-linked genes are in linkage dis-

equilibrium with those that control drive, potentially limit-

ing the possibility of compensatory testes enlargement and

explaining why T. whitei drive males have reduced fertility

under sperm competition. The second study of Wilkinson et

al. (2006) used a similar double mating design in T. dalmanni

(although only with SR males in the P2 role). As in this study,

it reported no difference between SR and ST success in mixed

paternity broods. However, in single-parent broods (where

only one male fathered offspring), there were 11 from the ST

male and only three from the SR male (rate 14/40 = 35%).

In this study, we found the pattern was reversed with three

from the ST male and eight from the SR male (rate 11/51 =

22%). There were experimental design differences that might

be important. In particular, Wilkinson et al. (2006) took ex-

perimental males from mixed sex cages with no control over

prior mating, whereas we kept males without females for

several days to allow their accessory glands to return to full

size (Rogers et al., 2005). This could explain the higher rate

of single-parent broods in Wilkinson et al. (2006). However,

combining across these two studies, we conclude that there

can be little condence that there is a large decit in SR male

single-parent broods. This is consistent with previous work,

which showed no difference in the failure rate of sperm trans-

fer to the spermatheca of females mated once either to ST or

SR males (Meade et al., 2019).

In line with earlier work on sperm competition in stalk-

eyed ies, there was no effect of mating order on paternity,

suggesting that the sperm of the rst and second male simply

mix and there is no sperm precedence in T. dalmanni (Bellamy,

2012; Corley et al., 2006; Wilkinson & Fry, 2001). Corley et

al. (2006) found evidence of a trimodal P2 distribution, cen-

tered around equal paternity as well as a strong bias to either

the rst or second male (double matings with ST males). This

contrasts with the at distribution shown here (Figure 1). The

difference could be due to the multiple mating design used by

Corley et al. (2006), in which each female was mated three

times with the rst and second males. A trimodal pattern was

also reported in a double mating design in the distantly relat-

ed South African stalk-eyed y species Diasemopsis meigenii,

where extreme paternity bias was explained by the failure

of sperm transfer after a single copulation (Bellamy, 2012).

Whatever the explanation, none of these studies support the

idea of a competitive advantage associated with mating posi-

tion in stalk-eyed ies.

The lack of difference found in this study may be limited by

sample size (n = 51), like all statistical comparisons. We ad-

dressed this by re-sampling the data with up to a 10-fold in-

ation in sample size. This increased the likelihood of nding

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2331Evolution (2023), Vol. 77, No. 10

a mating order difference (favoring P2 at a 4-fold increase in

sample size) to a much greater extent than a genotypic differ-

ence (no difference even at a 10-fold increase in sample size).

Given that these comparisons rely on the same distribution of

the data, they allow us to conclude that if there is a difference

in the paternity gain due to genotype, it is of a lower order

than that relating to mating order, and there is no evidence

to support the hypothesis of a competitive disadvantage as-

sociated with drive (if anything, the data favors a SR advan-

tage). Our approach is not wholly satisfactory as re-sampling

maintains the distribution of offspring genotyped per female,

which was variable (95% condence range 19–26), although

to some extent this is accounted for by the binomial tests.

A re-sampling of this distribution would inevitably require

further assumptions and end-up being contrived. We adopted

an approach that maintains the distribution of offspring gen-

otyped per female to frame our conclusions within the limita-

tions of the data collected.

In this study of T. dalmanni, sperm competition was as-

sessed under low-stress conditions. Virgin females were mat-

ed to two males separated by a 24-hr period. Experimental

males were not virgins but had been kept for several days

in single-sex groups. The objective was to assess SR and ST

males under standardized conditions as a rst step to under-

standing how SR males perform under sperm competition.

This is a highly specic experiment, designed to test whether

a male gains an advantage after a single competitive mating,

either because there is rst/last male precedence or variation

due to genotype. In the wild, competitive conditions are more

complex. Males form leks with multiple females at dusk

and then mate in a short period at dawn before dispersal,

with occasional matings interspersed during daylight hours

(Chapman et al., 2005; Cotton et al., 2010, 2015; Wilkinson

et al., 1998b). Females mate repeatedly in a life span that can

extend over several months (Reguera et al., 2004; Wilkinson

et al., 1998b). Multiple matings are required to maxise fer-

tility as males transfer low numbers of sperm per ejaculate

(Meade et al., 2019; Rogers et al., 2006; Wilkinson et al.,

2005; Baker, 2001), and sperm usage leads to a quick drop

in female fertility over time (Meade et al., 2017; Wilkinson et

al., 1998a). Future experiments need to assess the success of

single SR and ST male matings in females with a background

of multiple mating, closer to the conditions found in nature.

There may be differences when female sperm storage organs

are saturated compared to the situation with double mating

when females are below maximal fertility (Baker, 2001). In

addition, it will be important to assess the effect of the mating

rate, which is lower in SR males (Meade et al., 2020; Rogers

et al., 2008; Wilkinson et al., 2003). SR males may be less

able to compete in populations at high density where there

are multiple opportunities to mate, even though sperm trans-

fer does not differ between genotypes in sequential matings

over a short period of time (Meade et al., 2019). These fur-

ther studies will provide a more comprehensive assessment of

sperm competition as a factor contributing to the fertility of

drive males and its consequences for the frequency of SR in

wild populations.

In summary, we demonstrate that meiotic drive is not al-

ways associated with male fertility reduction under condi-

tions of sperm competition, even though drive destroys half

of carrier-male sperm. The lack of a fertility cost potentially

contributes to the relatively high frequency of meiotic drive

in T. dalmanni, which is around 20% in wild populations

(Cotton et al., 2014; Paczolt et al., 2017; Wilkinson et al.,

2003). This pattern is unlike other species where drive males

do poorly under sperm competition and the spread of drive is

reliant on a high frequency of monandrous matings (Courret

et al., 2019; Dyer and Hall, 2019; Price et al., 2008b). The

absence of a fertility cost is likely an evolved response to the

loss of sperm caused by meiotic drive, which is supported by

the nding in T. dalmanni that drive male testes are larger

at eclosion, have higher growth rates and are considerably

enlarged at maturity (Bradshaw et al., 2022; Meade et al.,

2020). We provide strong evidence against the consensus that

drive males are outperformed by non-drive males under sperm

competition—which suggests that other species should be in-

vestigated for evidence of mitigation of drive fertility costs.

Supplementary material

Supplementary material is available online at Evolution.

Data availability

The data that support the ndings of this study are openly

available in the Dryad at https://doi.org/10.5061/dryad.mkk-

wh713g.

Author contributions

S.B., L.M., and A.P. conceived the study. S.B. and L.M. car-

ried out experiments and S.B. collected the data. S.B., L.M.,

and A.P. analyzed the data. S.B. and A.P. wrote the paper. All

authors read, reviewed, and agreed on the submitted version.

Conict of interest: The authors declare no conict of interest.

Acknowledgments

We thank Rebecca Finley, Wendy Hart, and Matteo Mondani

for maintaining the y stocks and help with genotyping,

and Kevin Fowler for advice on experimental design. S.B.

is supported by a Studentship from the BBSRC LIDo DTP

(BB/M009513/1), A.P. is supported by funding from the

Engineering and Physical Sciences Research Council (EP/

F500351/1, EP/I017909/1), Natural Environment Research

Council (NE/R010579/1, NE/X009734/1) and Biotechnology

and Biological Sciences Research Council (BB/V003542/1).

References

Angelard, C., Montchamp-Moreau, C., & Joly, D. (2008). Female-driv-

en mechanisms, ejaculate size and quality contribute to the lower

fertility of sex-ratio distorter males in Drosophila simulans. BMC

Evolutionary Biology, 8, 326. https://doi.org/10.1186/1471-2148-

8-326

Atlan, A., Joly, D., Capillon, C., & Montchamp-Moreau, C. (2004).

Sex-ratio distorter of Drosophila simulans reduces male

productivity and sperm competition ability. Journal of Evolu-

tionary Biology, 17(4), 744–751. https://doi.org/10.1111/j.1420-

9101.2004.00737.x

Baker, R. H. (2001). Effects of multiple mating and male eye span

on female reproductive output in the stalk-eyed y, Cyrtodiop-

sis dalmanni. Behavioral Ecology, 12(6), 732–739. https://doi.

org/10.1093/beheco/12.6.732

Baker, R. H., Denniff, M., Futerman, P., Fowler, K., Pomiankowski,

A., & Chapman, T. (2003). Accessory gland size inuences time to

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2332 Bates et al.

sexual maturity and mating frequency in the stalk-eyed y, Cyrto-

diopsis dalmanni. Behavioral Ecology, 14, 607–611.

Bellamy, L. A. R. (2012). Sexual selection in stalk-eyed ies; inbreeding

depression, sperm competition and larval development. UCL (Uni-

versity College London).

Bradshaw, S. L., Meade, L., Tarlton-Weatherall, J., & Pomiankowski,

A. (2022). Meiotic drive adaptive testes enlargement during early

development in the stalk-eyed y. Biology Letters 18, 20220352.

https://doi.org/10.1098/rsbl.2022.0352

Chapman, T., Pomiankowski, A., & Fowler, K. (2005). Stalk-eyed ies.

Current Biology, 15(14), R533–R535. https://doi.org/10.1016/j.

cub.2005.07.015

Burke, T. A., Bruford, M. W., Hanotte, O., & Brookeld, J. F. Y. (1998).

Multilocus and single-locus DNA ngerprinting. IRL Press.

Christianson, S. J., Brand, C. L., & Wilkinson, G. S. (2011). Reduced

polymorphism associated with x chromosome meiotic drive in

the stalk-eyed y Teleopsis dalmanni. PLoS One, 6(11), e27254.

https://doi.org/10.1371/journal.pone.0027254

Corley, L. S., Cotton, S., McConnell, E., Chapman, T., Fowler, K., &

Pomiankowski, A. (2006). Highly variable sperm precedence in the

stalk-eyed y, Teleopsis dalmanni. BMC Evolutionary Biology, 6,

53. https://doi.org/10.1186/1471-2148-6-53

Cotton, A. J., Cotton, S., Small, J., & Pomiankowski, A. (2015). Male

mate preference for female eyespan and fecundity in the stalk-

eyed y, Teleopsis dalmanni. Behavioral Ecology, 26(2), 376–385.

https://doi.org/10.1093/beheco/aru192

Cotton, A. J., Földvári, M., Cotton, S., & Pomiankowski, A. (2014).

Male eyespan size is associated with meiotic drive in wild stalk-

eyed ies (Teleopsis dalmanni). Heredity, 112(4), 363–369. https://

doi.org/10.1038/hdy.2013.131

Cotton, S., Small, J., Hashim, R., & Pomiankowski, A. (2010). Eyespan

reects reproductive quality in wild stalk-eyed ies. Evolutionary

Ecology, 24(1), 83–95. https://doi.org/10.1007/s10682-009-9292-

Courret, C., Chang, C. -H., Wei, K. H. -C., Montchamp-Moreau, C., &

Larracuente, A. M. (2019). Meiotic drive mechanisms: lessons from

Drosophila. Proceedings of the Royal Society B: Biological Scienc-

es, 286(1913), 20191430. https://doi.org/10.1098/rspb.2019.1430

David, P., Bjorksten, T., Fowler, K., & Pomiankowski, A. (2000). Con-

dition-dependent signalling of genetic variation in stalk-eyed ies.

Nature, 406(6792), 186–188. https://doi.org/10.1038/35018079

Dobler, R., Charette, M., Kaplan, K., Turnell, B. R., & Reinhardt, K.

(2022). Divergent natural selection alters male sperm competition

success in Drosophila melanogaster. Ecology and Evolution, 12(2),

e8567. https://doi.org/10.1002/ece3.8567

Dyer, K. A., & Hall, D. W. (2019). Fitness consequences of a non-re-

combining sex-ratio drive chromosome can explain its prevalence

in the wild. Proceedings of the Royal Society B: Biological Sciences,

286(1917), 20192529. https://doi.org/10.1098/rspb.2019.2529

Finnegan, S. R., White, N. J., Koh, D., Camus, F. M., Fowler, K., & Po-

miankowski, A. (2019). Meiotic drive reduces egg-to-adult viability

in stalk-eyed ies. Proceedings of the Royal Society B: Biological

Sciences, 286, 20191414. https://doi.org/10.1098/rspb.2019.1414

Gershenson, S. (1928). A new sex-ratio abnormality in Drosophila

obscura. Genetics, 13(6), 488–507. https://doi.org/10.1093/genet-

ics/13.6.488

Haig, D., & Bergstrom, C. T. (1995). Multiple mating, sperm compe-

tition and meiotic drive. Journal of Evolutionary Biology, 8(3),

265–282. https://doi.org/10.1046/j.1420-9101.1995.8030265.x

Hamilton, W. D. (1967). Extraordinary sex ratios. Science, 156(3774),

477–488. https://doi.org/10.1126/science.156.3774.477

Hatcher, M. J., Taneyhill, D. E., Dunn, A. M., & Tofts, C. (1999).

Population dynamics under parasitic sex ratio distortion. Theo-

retical Population Biology, 56(1), 11–28. https://doi.org/10.1006/

tpbi.1998.1410

Hingle, A., Fowler, K., & Pomiankowski, A. (2001). Size-depen-

dent mate preference in the stalk-eyed y Cyrtodiopsis dalman-

ni. Animal Behaviour, 61(3), 589–595. https://doi.org/10.1006/

anbe.2000.1613

Howie, J. M., Dawson, H. A. C., Pomiankowski, A., & Fowler, K.

(2019). Limits to environmental masking of genetic quality in sexu-

al signals. Journal of Evolutionary Biology, 32(8), 868–877. https://

doi.org/10.1111/jeb.13491

Hurst, L. D., & Pomiankowski, A. (1991). Causes of sex ratio bias may

account for unisexual sterility in hybrids: A new explanation of

Haldane’s rule and related phenomena. Genetics, 128(4), 841–858.

https://doi.org/10.1093/genetics/128.4.841

James, A. C., & Jaenike, J. (1990). “Sex ratio” meiotic drive in Drosoph-

ila testacea. Genetics, 126(3), 651–656. https://doi.org/10.1093/

genetics/126.3.651

Jiggins, F. M., Hurst, G. D. D., & Majerus, M. E. N. (1999). How com-

mon are meiotically driving sex chromosomes in insects? American

Naturalist, 154(4), 481–483. https://doi.org/10.1086/303251

Kotrba, M. (1995). The internal female genital organs of Chaetodiopsis

and Diasemopsis (Diptera: Diopsidae) and their systematic rele-

vance. Annals of the Natal Museum, 36, 147–159.

Lindholm, A. K., Dyer, K. A., Firman, R. C., Fishman, L., Forstmeier,

W., Holman, L., Johannesson, H., Knief, U., Kokko, H., Larrac-

uente, A. M., Manser, A., Montchamp-Moreau, C., Petrosyan, V.

G., Pomiankowski, A., Presgraves, D. C., Safronova, L. D., Sutter,

A., Unckless, R. L., Verspoor, R. L., … Price, T. A. R. (2016). The

ecology and evolutionary dynamics of meiotic drive. Trends in

Ecology and Evolution, 31(4), 315–326. https://doi.org/10.1016/j.

tree.2016.02.001

Lorch, P. D., Wilkinson, G. S., & Reillo, P. R. (1993). Copulation du-

ration and sperm precedence in the stalk-eyed y Cyrtodiopsis

whitei (Diptera: Diopsidae). Behavior, Ecology and Sociobiology,

32, 303–311.

Mackintosh, C., Pomiankowski, A., & Scott, M. F. (2021). X-linked

meiotic drive can boost population size and persistence. Genetics,

217(1), 1–11. https://doi.org/10.1093/genetics/iyaa018

Meade, L., Harley, E., Cotton, A., Howie, J. M., Pomiankowski, A.,

& Fowler, K. (2017). Variation in the benets of multiple mating

on female fertility in wild stalk-eyed ies. Ecology and Evolution,

7(23), 10103–10115. https://doi.org/10.1002/ece3.3486

Meade, L. C., Dinneen, D., Kad, R., Lynch, D. M., Fowler, K., & Po-

miankowski, A. (2019). Ejaculate sperm number compensation in

stalk-eyed ies carrying a selsh meiotic drive element. Heredity,

122(6), 916–926. https://doi.org/10.1038/s41437-018-0166-y

Meade, L. C., Finnegan, S. R., Kad, R., Fowler, K., & Pomiankowski, A.

(2020). Maintenance of fertility in the face of meiotic drive. Amer-

ican Naturalist, 195, 743–751.

Meier, R., & Baker, R. H. (2002). A cladistic analysis of Diopsidae

(Diptera) based on morphological and DNA sequence data. Insect

Systematics and Evolution, 33, 325–336.

Newton, M. E., Wood, R. J., & Southern, D. I. (1976). A cytogenetic

analysis of meiotic drive in the mosquito, Aedes aegypti (L.). Ge-

netica, 46(3), 297–318. https://doi.org/10.1007/bf00055473

Paczolt, K. A., Reinhardt, J. A., & Wilkinson, G. S. (2017). Contrasting

patterns of X-chromosome divergence underlie multiple sex-ratio

polymorphisms in stalk-eyed ies. Journal of Evolutionary Biology,

30(9), 1772–1784. https://doi.org/10.1111/jeb.13140

Parker, G. A. (1970). Sperm competition and its evolutionary conse-

quences in the insects. Biological Review, 45(4), 525–567. https://

doi.org/10.1111/j.1469-185x.1970.tb01176.x

Policansky, D. (1974). “Sex ratio,” meiotic drive, and group selection in

Drosophila pseudoobscura. American Naturalist, 108(959), 75–90.

https://doi.org/10.1086/282886

Presgraves, D. C., Baker, R. H., & Wilkinson, G. S. (1999). Coevo-

lution of sperm and female reproductive tract morphology in

stalk–eyed ies. Proceedings of the Royal Society London B: Bi-

ological Sciences, 266(1423), 1041–1047. https://doi.org/10.1098/

rspb.1999.0741

Presgraves, D. C., Severance, E., & Wilkinson, G. S. (1997). Sex chro-

mosome meiotic drive in stalk-eyed ies. Genetics, 147(3), 1169–

1180. https://doi.org/10.1093/genetics/147.3.1169

Price, T. A. R., Bretman, A. J., Avent, T. D., Snook, R. R., Hurst, G.

D. D., & Wedell, N. (2008a). Sex ratio distorter reduces sperm

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

2333Evolution (2023), Vol. 77, No. 10

competitive ability in an insect. Evolution, 62(7), 1644–1652.

https://doi.org/10.1111/j.1558-5646.2008.00386.x

Price, T. A. R., Hodgson, D. J., Lewis, Z., Hurst, G. D. D., & Wedell,

N. (2008b). Selsh genetic elements promote polyandry in a y.

Science, 322(5905), 1241–1243. https://doi.org/10.1126/sci-

ence.1163766

Price, T. A. R., & Wedell, N. (2008). Selsh genetic elements and sexual

selection: their impact on male fertility. Genetica, 134(1), 99–111.

https://doi.org/10.1007/s10709-008-9253-y

Reguera, P., Pomiankowski, A., Fowler, K., & Chapman, T. (2004).

Low cost of reproduction in female stalk-eyed ies, Cyrtodiopsis

dalmanni. Journal of Insect Physiology, 50(1), 103–108. https://

doi.org/10.1016/j.jinsphys.2003.10.004

R Core Team. (2021). R: A language and environment for statisti-

cal computing. R Foundation for Statistical Computing, Vienna,

Austria. https://www.R-project.org/

Reinhardt, J. A., Baker, R. H., Zimin, A. V., Ladias, C., Paczolt, K. A.,

Werren, J. H., Hayashi, C. Y., & Wilkinson, G. S. (2023). Impacts of

sex ratio meiotic drive on genome structure and function in a stalk-

eyed y. Genome Biology and Evolution, 15(7), evad118. https://

doi.org/10.1093/gbe/evad118

Reinhardt, J. A., Brand, C. L., Paczolt, K. A., Johns, P. M., Baker, R.

H., & Wilkinson, G. S. (2014). Meiotic drive impacts expres-

sion and evolution of X-linked genes in stalk-eyed ies. PLoS

Genetics, 10(5), e1004362. https://doi.org/10.1371/journal.

pgen.1004362

Rogers, D. W., Chapman, T., Fowler, K., & Pomiankowski, A. (2005).

Mating-induced reduction in accessory reproductive organ size in

the stalk-eyed y, Cyrtodiopsis dalmanni. BMC Evolutionary Biol-

ogy, 5, 37. https://doi.org/10.1186/1471-2148-5-37

Rogers, D. W., Denniff, M., Chapman, T., Fowler, K., & Pomiankowski,

A. (2008). Male sexual ornament size is positively associated with

reproductive morphology and enhanced fertility in the stalk-eyed

y, Teleopsis dalmanni. BMC Evolutionary Biology, 8, 236. https://

doi.org/10.1186/1471-2148-8-236

Rogers, D. W., Grant, C. A., Chapman, T., Pomiankowski, A., & Fowler,

K. (2006). The inuence of male and female eyespan on fertility in

the stalk-eyed y, Cyrtodiopsis dalmanni. Animal Behaviour, 72(6),

1363–1369. https://doi.org/10.1016/j.anbehav.2006.03.027

Sandler, L., & Novitski, E. (1957). Meiotic drive as an evolution-

ary force. American Naturalist, 91(857), 105–110. https://doi.

org/10.1086/281969

Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Im-

age to ImageJ: 25 years of image analysis. Nature Methods, 9(7),

671–675. https://doi.org/10.1038/nmeth.2089

Searle, J. B., & de Villena, F. P-M. (2022). The evolutionary signicance

of meiotic drive. Heredity, 129, 44–47.

Verspoor, R. L., Price, T. A. R., & Wedell, N. (2020). Selsh genetic

elements and male fertility. Philosophical Transaction Royal So-

ciety B: Biological Sciences, 375(1813), 20200067. https://doi.

org/10.1098/rstb.2020.0067

Wilkinson, G. S., Amitin, E. G., & Johns, P. M. (2005). Sex-linked cor-

related responses in female reproductive traits to selection on male

eye span in stalk-eyed ies. Integrative and Comparative Biology,

45(3), 500–510. https://doi.org/10.1093/icb/45.3.500

Wilkinson, G. S., & Fry, C. L. (2001). Meiotic drive alters sperm com-

petitive ability in stalk-eyed ies. Proceedings of the Royal Soci-

ety B: Biological Sciences, 268(1485), 2559–2564. https://doi.

org/10.1098/rspb.2001.1831

Wilkinson, G. S., Johns, P. M., Kelleher, E. S., Muscedere, M. L., &

Lorsong, A. (2006). Fitness effects of X chromosome drive in

the stalk-eyed y, Cyrtodiopsis dalmanni. Journal of Evolution-

ary Biology, 19(6), 1851–1860. https://doi.org/10.1111/j.1420-

9101.2006.01169.x

Wilkinson, G. S., Kahler, H., & Baker, R. H. (1998a). Evolution of fe-

male mating preferences in stalk-eyed ies. Behavioral Ecology,

9(5), 525–533. https://doi.org/10.1093/beheco/9.5.525

Wilkinson, G. S., Presgraves, D. C., & Crymes, L. (1998b). Male eye span

in stalk-eyed ies indicates genetic quality by meiotic drive suppres-

sion. Nature, 391(6664), 276–279. https://doi.org/10.1038/34640

Wilkinson, G. S., Swallow, J. G., Christensen, S. J., & Madden, K.

(2003). Phylogeography of sex ratio and multiple mating in stalk-

eyed ies from southeast Asia. Genetica, 117, 37–46.

Wolf, J. B., Ferguson-Smith, A. C., & Lorenz, A. (2022). Mendel’s laws

of heredity on his 200th birthday: What have we learned by consid-

ering exceptions? Heredity, 129(1), 1–3. https://doi.org/10.1038/

s41437-022-00552-y

Zanders, S. E., & Unckless, R. L. (2019). Fertility costs of meiot-

ic drivers. Current Biology, 29(11), R512–R520. https://doi.

org/10.1016/j.cub.2019.03.046

Zeh, J. A., & Zeh, D. W. (1997). The evolution of polyandry II: post–

copulatory defenses against genetic incompatibility. Proceedings

of the Royal Society London B: Biological Sciences, 264(1378),

69–75. https://doi.org/10.1098/rspb.1997.0010

Downloaded from https://academic.oup.com/evolut/article/77/10/2326/7249914 by guest on 03 November 2023

The following pages contain Supplementary Information from: “Meiotic drive does not

impede success in sperm competition in the stalk-eyed fly, Teleopsis dalmanni”.

Supplementary Figures have been renumbered in this document to fit the main thesis

text.

Authors: Sadé Bates1, Lara Meade1 and Andrew Pomiankowski1,2

1 Department of Genetics, Evolution and Environment, University College London, Gower

Street, London, WC1E 6BT, UK

2 CoMPLEX, University College London, Gower Street, London, WC1E 6BT, UK

Address correspondence to: A. Pomiankowski. E-mail: ucbhpom@ucl.ac.uk

ORCID

Sade Bates, 0000-0002-9736-1077

Lara Meade, 0000-0002-5724-7413

Andrew Pomiankowski, 0000-0002-5171-8755

Supplementary Information 1:

• supplementary methods

Supplementary Information 2:

• all statistical models and effect sizes for all broods

• supplementary Figures S2.2.1 and S2.2.2

Supplementary Information 3:

• all statistical models and effect sizes for broods with ≥ 10 genotyped offspring

Supplementary Information 4:

• resampling

Supplementary Information 1: supplementary methods

The following protocol was used to extract and purify larval DNA (adapted from standard

protocol in Burke et al., 1998):

Larvae within each well were crushed using a micro-pestle prior to incubation for 16hrs at

55ºC, to extract DNA. The following day, 35µL 4M ammonium acetate was added to each

sample to precipitate out proteins, and the plates chilled on ice for 5mins. The plates were

then spun at 4450rpm, 4ºC for 60min. Next, the DNA was precipitated out by transferring

80µL of the supernatant from each sample to a new plate containing 80µL isopropanol per

well. Centrifugation at 4450rpm, 4ºC for 60min pelleted out the DNA. The supernatant was

discarded, and the DNA pellets were washed by adding 100µL 70% ethanol and spinning at

4450rpm for 30min. The ethanol was then removed, and the plates left to air dry for 1hr

before adding 30µL T10 E0.1 buffer to each sample and incubating at 37ºC for 30mins to re-

dissolve the DNA. Samples were stored at -20ºC prior to PCR analysis.

References

Burke, T. A. et al. (1998) ‘Multilocus and single-locus DNA fingerprinting’. IRL Press.

The following PCR conditions were used for progeny genotyping:

A 2720 Thermal Cycler (Applied Biosystems, Woolston, UK) was used to perform the

reactions, which were carried out in 11µL volumes per sample, containing: 0.6µL forward

and 0.6µL reverse primers (see Supp. Table 1 for sequences), both at 10µM, 0.12µL

Phusion® High-Fidelity DNA Polymerase (New England Biolabs, Herts), 2.4µL Phusion® HF

buffer (New England Biolabs, Herts), 6.4µL ddH20 and either 1µL 10x diluted DNA, 5x

diluted or pure DNA (depending on DNA concentration). The PCR programme was a 10min

initial denaturation stage at 98ºC, followed by 45 cycles of 10sec denaturation at 98ºC, 30sec

annealing time at 63ºC and 20sec extension at 72ºC. The reaction was completed by a 7min

final extension step at 72ºC. The PCR products were analysed via gel electrophoresis on a 3%

agarose/TBE gel run at 100V for ~1hr to separate them according to size and the results were

visualised using a gel imaging system.

Supplementary Table 1: comp16710 primer sequences

STRAND

Sequence

Forward

CGTGTCCGCATTTATACCAC

Reverse

GGTAGGCTTGTTCTAACGGC

Supplementary Information 2: all model tables and eﬀect sizes for

all broods

Contents

1 Data 3

2 Male fertility 3

2.1 Variation in number of oﬀspring sired with mating position . . . . . . . . . . . . . . . . . . . 3

2.2 Variation in number of oﬀspring sired with male genotype . . . . . . . . . . . . . . . . . . . . 3

2.3 Variation in number of oﬀspring sired with male genotype in the P1 role or P2 role . . . . . . 4

2.4 Number of larvae collected and batch number . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.4.1 Variation in P2 number of oﬀspring sired with larvae collected . . . . . . . . . . . . . 5

2.4.2 Variation in SR number of oﬀspring sired with larvae collected . . . . . . . . . . . . . 5

2.4.3 Variation in P2 number of oﬀspring sired with larvae collected and mating order . . . 6

2.4.4 Variation in SR number of oﬀspring sired with larvae collected and genotype . . . . . 6

2.4.5 Variation in P2 number of oﬀspring sired with mating order and batch number . . . . 7

2.4.6 Variation in SR number of oﬀspring sired with genotype and batch number . . . . . . 7

2.5 Male fertility and single parent broods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.5.1 Variation in number of oﬀspring sired with mating position with only single parent

broods............................................. 8

2.5.2 Variation in number of oﬀspring sired with male genotype with only single parent broods 8

2.5.3 Variation in number of oﬀspring sired with mating position when single parent broods

areexcluded ......................................... 9

2.5.4 Variation in male number of oﬀspring sired with male genotype when single parent

broodsareexcluded ..................................... 10

3 Male trait size and mating duration 10

3.1 Variationinmaletraits ....................................... 10

3.2 Male fertility with thorax length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.3 Male fertility with mating duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.3.1 Variation in mating duration with mating position . . . . . . . . . . . . . . . . . . . . 13

3.3.2 Variation in mating duration with male genotype . . . . . . . . . . . . . . . . . . . . . 14

3.3.3 The eﬀect of mating duration on the number of oﬀspring sired per mating position . . 14

3.3.4 The eﬀect of mating duration on the number of oﬀspring sired per male genotype . . . 15

3.3.5 The eﬀect of mating position with mating duration on the number of oﬀspring sired . 15

3.3.6 The eﬀect of male genotype with mating duration on the number of oﬀspring sired . . 16

4 Supplementary ﬁgures 17

1 Data

This analysis includes all broods with more than 2 oﬀspring genotyped, N = 51. 2 matings were 29secs in

duration, however these copulations were still deemed successful they resulted in oﬀspring.

1 ST male failed to mate in the P1 position and was replaced, and 1 SR male failed to mate in the in P2

position and was replaced. 1 female failed to mate with any male and was discarded.

2 Male fertility

2.1 Variation in number of oﬀspring sired with mating position

glm(formula = cbind(SR_offspring, ST_offspring) ~ as.factor(mating_position),

family = quasibinomial, data = by_brood2)

Table 1: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 50 554.310

as.factor(mating_position) 1 12.022 49 542.288 1.307 0.259

Table 2: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.125 0.250 -0.501 0.619

as.factor(mating_position)P2 0.409 0.358 1.140 0.260

The mean proportion of oﬀspring sired by the P2 male was 0.522 ±0.327 (mean P2 ±sd) and there was no

eﬀect of mating position on the number of oﬀspring sired per male.

2.2 Variation in number of oﬀspring sired with male genotype

glm(formula = cbind(P2_offspring, P1_offspring) ~ as.factor(male_genotype),

family = quasibinomial, data = by_brood2)

Table 3: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 50 544.092

as.factor(male_genotype) 1 1.805 49 542.288 0.196 0.66

Table 4: Model coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.284 0.257 1.103 0.275

as.factor(male_genotype)ST -0.159 0.358 -0.443 0.660

The mean proportion of oﬀspring sired by the SR male was 0.577 ±0.319 (SR ±sd) and there was no eﬀect

of male genotype on the number of oﬀspring sired per male.

2.3 Variation in number of oﬀspring sired with male genotype in the P1 role

or P2 role

lm(formula = P1_offspring ~ as.factor(male_genotype), data = P1_males_only)

Table 5: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

as.factor(male_genotype) 1 0.214 0.214 0.002 0.963

Residuals 49 4916.963 100.346

Table 6: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 10.296 1.928 5.341 0.000

as.factor(male_genotype)ST -0.130 2.810 -0.046 0.963

lm(formula = P2_offspring ~ as.factor(male_genotype), data = P2_males_only)

Table 7: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

as.factor(male_genotype) 1 42.706 42.706 0.434 0.513

Residuals 49 4826.000 98.490

Table 8: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 13.500 2.026 6.664 0.000

as.factor(male_genotype)ST -1.833 2.784 -0.658 0.513

Examining the data from P1 males only, there was no eﬀect of genotype on number of oﬀspring sired by the

P1 male (F1,49 = 0.002, P= 0.963). The same was true when examining data from P2 males only (F1,49 =

0.434, P= 0.513).

2.4 Number of larvae collected and batch number

The number of larvae collected is a measure of female fecundity, as it is the total number of oﬀspring that

were collected including the random sample of oﬀspring from each female weren’t genotyped (for logistic

reasons). The mean number of larvae collected per female was 48.196 ±22.735 (mean ±sd), with a range

of 6 - 116 oﬀspring.

2.4.1 Variation in P2 number of oﬀspring sired with larvae collected

glm(formula = cbind(P2_offspring, P1_offspring) ~ larvae_collected,

family = quasibinomial, data = by_brood2)

Table 9: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 50 544.092

larvae_collected 1 0.342 49 543.750 0.037 0.848

Table 10: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.287 0.475 0.605 0.548

larvae_collected -0.001 0.008 -0.193 0.848

2.4.2 Variation in SR number of oﬀspring sired with larvae collected

glm(formula = cbind(SR_offspring, ST_offspring) ~ larvae_collected,

family = quasibinomial, data = by_brood2)

Table 11: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 50 554.310

larvae_collected 1 44.381 49 509.928 5.09 0.029

Table 12: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 1.060 0.481 2.201 0.032

larvae_collected -0.017 0.008 -2.193 0.033

The number of Larvae collected was not associated with the numbers of P2 or SR oﬀspring.

2.4.3 Variation in P2 number of oﬀspring sired with larvae collected and mating order

glm(formula = cbind(SR_offspring, ST_offspring) ~ larvae_collected +

mating_position, family = quasibinomial, data = by_brood2)

Table 13: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

larvae_collected 58.453 1 6.745 0.012

mating_position 26.093 1 3.011 0.089

Residuals 415.988 48

Table 14: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.939 0.488 1.925 0.060

larvae_collected -0.020 0.008 -2.514 0.015

mating_positionP2 0.634 0.369 1.716 0.093

2.4.4 Variation in SR number of oﬀspring sired with larvae collected and genotype

glm(formula = cbind(P2_offspring, P1_offspring) ~ larvae_collected +

male_genotype, family = quasibinomial, data = by_brood2)

Table 15: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

larvae_collected 0.805 1 0.086 0.771

male_genotype 2.268 1 0.242 0.625

Residuals 450.489 48

Table 16: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.429 0.561 0.765 0.448

larvae_collected -0.002 0.008 -0.293 0.771

male_genotypeST -0.182 0.371 -0.491 0.626

When the number of larvae collected was added as a covariate, there was no aﬀect on the trends previously

observed: neither mating order nor male genotype had an aﬀect on the number of oﬀspring sired by each

male (P> 0.05).

2.4.5 Variation in P2 number of oﬀspring sired with mating order and batch number

glm(formula = cbind(SR_offspring, ST_offspring) ~ batch + mating_position,

family = quasibinomial, data = by_brood2)

Table 17: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

batch 26.229 1 2.881 0.096

mating_position 12.733 1 1.399 0.243

Residuals 437.002 48

Table 18: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -1.134 0.652 -1.740 0.088

batch 0.623 0.369 1.687 0.098

mating_positionP2 0.425 0.361 1.179 0.244

2.4.6 Variation in SR number of oﬀspring sired with genotype and batch number

glm(formula = cbind(P2_offspring, P1_offspring) ~ batch + male_genotype,

family = quasibinomial, data = by_brood2)

Table 19: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

batch 14.562 1 1.581 0.215

male_genotype 1.954 1 0.212 0.647

Residuals 442.084 48

Table 20: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.453 0.640 -0.708 0.482

batch 0.462 0.368 1.255 0.216

male_genotypeST -0.166 0.361 -0.460 0.647

The eﬀects of mating order and genotype remained insigniﬁcant (P> 0.05) when Batch number (mating

trials were carried out in two batches, for ease) was added as a covariate.

2.5 Male fertility and single parent broods

Table 21: Single parent broods

mating position proportion P2 oﬀspring male genotype proportion SR oﬀspring

P1 0.028 ST 0.972

P1 0.000 ST 1.000

P1 0.956 ST 0.044

P1 0.000 ST 1.000

P2 1.000 SR 1.000

P1 0.000 ST 1.000

P2 0.955 SR 0.955

P2 0.958 SR 0.958

P1 0.000 ST 1.000

P1 0.967 ST 0.033

P1 0.000 ST 1.000

P2 0.960 SR 0.960

P2 0.969 SR 0.969

P2 0.034 SR 0.034

P2 1.000 SR 1.000

2.5.1 Variation in number of oﬀspring sired with mating position with only single parent

broods

glm(formula = cbind(SR_offspring, ST_offspring) ~ as.factor(mating_position),

family = quasibinomial, data = subset(by_brood2, proportion_SR <=

0.05 | proportion_SR >= 0.95))

Table 22: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 14 325.901

as.factor(mating_position) 1 24.816 13 301.085 1.188 0.296

Table 23: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.152 0.729 0.209 0.838

as.factor(mating_position)P2 1.226 1.162 1.056 0.310

When only extreme P2 oﬀspring proportions of ≤0.05 and ≥0.95 are considered, the proportion of oﬀspring

sired by the P2 male was not diﬀerent from 0.5 (proportion P2 = 0.522 ±0.497), and mating order had no

eﬀect on the number of oﬀspring sired.

2.5.2 Variation in number of oﬀspring sired with male genotype with only single parent

broods

glm(formula = cbind(P2_offspring, P1_offspring) ~ as.factor(male_genotype),

family = quasibinomial, data = subset(by_brood2, proportion_P2 <=

0.05 | proportion_P2 >= 0.95))

Table 24: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 14 340.750

as.factor(male_genotype) 1 39.665 13 301.085 1.898 0.192

Table 25: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 1.378 0.904 1.525 0.151

as.factor(male_genotype)ST -1.531 1.162 -1.318 0.210

When only extreme SR oﬀspring proportions of ≤0.05 and ≥0.95 are considered, the proportion of oﬀspring

sired by the SR male did not diﬀer from 0.5 (proportion SR ±sd = 0.795 ±0.393), and male genotype had

no eﬀect on number of oﬀspring sired.

2.5.3 Variation in number of oﬀspring sired with mating position when single parent broods

are excluded

glm(formula = cbind(SR_offspring, ST_offspring) ~ as.factor(mating_position),

family = quasibinomial, data = subset(by_brood2, proportion_SR >

0.05 & proportion_SR < 0.95))

Table 26: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 35 188.308

as.factor(mating_position) 1 1.224 34 187.084 0.243 0.625

Table 27: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.226 0.216 -1.045 0.303

as.factor(mating_position)P2 0.153 0.310 0.493 0.625

When extreme P2 oﬀspring proportions of ≤0.05 and ≥0.95 are excluded, the proportion of oﬀspring sired

by the P2 male was not diﬀerent from 0.5 (proportion P2 ±sd = 0.522 ±0.233) and mating order had no

eﬀect on number of oﬀspring sired.

2.5.4 Variation in male number of oﬀspring sired with male genotype when single parent

broods are excluded

glm(formula = cbind(P2_offspring, P1_offspring) ~ as.factor(male_genotype),

family = quasibinomial, data = subset(by_brood2, proportion_P2 >=

0.05 & proportion_P2 <= 0.95))

Table 28: Analysis of variance table

Df Deviance Resid. Df Resid. Dev F Pr(>F)

NULL 35 191.796

as.factor(male_genotype) 1 4.711 34 187.084 0.936 0.34

Table 29: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.073 0.222 -0.331 0.743

as.factor(male_genotype)ST 0.300 0.310 0.966 0.341

When extreme SR oﬀspring proportions of ≤0.05 and ≥0.95 are excluded, the proportion of oﬀspring sired

by the SR male was not diﬀerent from 0.5 (proportion SR ±sd = 0.486 ±0.234) and male genotype had no

eﬀect on number of oﬀspring sired.

3 Male trait size and mating duration

3.1 Variation in male traits

Table 30: Mean thorax length per male genotype

male_genotype N thorax sd se ci

SR 49 2.190 0.163 0.023 0.047

ST 50 2.297 0.178 0.025 0.051

Table 31: Mean eyespan per male genotype

male_genotype N eyespan sd se ci

SR 49 7.304 0.776 0.111 0.223

ST 50 7.897 0.815 0.115 0.232

Table 32: Mean residual eyespan per male genotype

male_genotype N residual_eyespan sd se ci

SR 49 -0.112 0.478 0.068 0.137

ST 50 0.092 0.531 0.075 0.151

lm(formula = eyespan ~ thorax, data = by_male_id2)

Table 33: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

thorax 1 44.406 44.406 167.242 0

Residuals 97 25.756 0.266

Table 34: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.875 0.658 -1.330 0.187

thorax 3.778 0.292 12.932 0.000

lm(formula = thorax ~ male_genotype, data = by_male_id2)

Table 35: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

male_genotype 1 0.285 0.285 9.783 0.002

Residuals 97 2.826 0.029

Table 36: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 2.190 0.024 89.812 0.000

male_genotypeST 0.107 0.034 3.128 0.002

lm(formula = eyespan ~ male_genotype, data = by_male_id2)

Table 37: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

male_genotype 1 8.720 8.720 13.766 0

Residuals 97 61.442 0.633

Table 38: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 7.304 0.114 64.239 0

male_genotypeST 0.594 0.160 3.710 0

lm(formula = eyespan ~ thorax + male_genotype, data = by_male_id2)

Table 39: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

thorax 36.651 1 141.924 0.000

male_genotype 0.964 1 3.734 0.056

Residuals 24.791 96

Table 40: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -0.583 0.666 -0.875 0.384

thorax 3.601 0.302 11.913 0.000

male_genotypeST 0.207 0.107 1.932 0.056

Thorax length varies between male genotypes. Eyespan has strong covariance with thorax. Relative eyespan

— the variation in eyespan not predicted by thorax length — is not signiﬁcantly diﬀerent between genotypes.

Therefore, relative eyespan is not added to binomial GLMs for paternity.

3.2 Male fertility with thorax length

glm(formula = cbind(ST_offspring, SR_offspring) ~ thorax.P1 +

thorax.P2 + mating_position, family = quasibinomial, data = by_brood2)

Table 41: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

thorax.P1 9.464 1 0.977 0.328

thorax.P2 0.515 1 0.053 0.819

mating_position 11.241 1 1.161 0.287

Residuals 426.130 44

Table 42: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 3.139 3.542 0.886 0.380

thorax.P1 -1.058 1.076 -0.984 0.330

thorax.P2 -0.291 1.265 -0.230 0.819

mating_positionP2 -0.452 0.421 -1.074 0.289

glm(formula = cbind(P1_offspring, P2_offspring) ~ thorax.ST +

thorax.SR + male_genotype, family = quasibinomial, data = by_brood2)

Table 43: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

thorax.ST 0.023 1 0.002 0.961

thorax.SR 7.167 1 0.731 0.397

male_genotype 3.619 1 0.369 0.547

Residuals 431.689 44

Table 44: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -2.510 3.512 -0.715 0.479

thorax.ST -0.055 1.127 -0.049 0.961

thorax.SR 1.026 1.204 0.852 0.399

male_genotypeST 0.236 0.388 0.607 0.547

Though thorax length is diﬀerent between SR and ST males, it does not aﬀect number of oﬀspring sired by

each male, nor does it aﬀect number of oﬀspring sired by P2 and SR males.

3.3 Male fertility with mating duration

Mating duration is the observed time taken for a single copulation in seconds.

3.3.1 Variation in mating duration with mating position

lm(formula = mating_duration_sec ~ mating_position, data = by_male_id2)

Table 45: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

mating_position 1 2306.127 2306.127 0.943 0.334

Residuals 100 244633.451 2446.335

Table 46: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 63.941 6.926 9.232 0.000

mating_positionP2 9.510 9.795 0.971 0.334

Mating duration did not diﬀer between P1 and P2 males (mean P1 mating duration ±se: 63.94sec ±3.43sec,

mean P2 mating duration ±se: 73.45sec ±3.43sec; F1,100 = 0.943, P= 0.334).

3.3.2 Variation in mating duration with male genotype

lm(formula = mating_duration_sec ~ male_genotype, data = by_male_id2)

Table 47: Analysis of variance table

Df Sum Sq Mean Sq F value Pr(>F)

male_genotype 1 6321.657 6321.657 2.627 0.108

Residuals 100 240617.922 2406.179

Table 48: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 76.569 6.869 11.147 0.000

male_genotypeST -15.745 9.714 -1.621 0.108

Mating duration did not diﬀer between ST and SR males (mean ST mating duration ±se: 60.82sec ±

2.36sec, mean SR mating duration ±se: 76.57sec ±9.42sec; F1,100 = 2.627, P= 0.108).

3.3.3 The eﬀect of mating duration on the number of oﬀspring sired per mating position

glm(formula = cbind(P2_offspring, P1_offspring) ~ mating_duration_sec.P1 +

mating_duration_sec.P2, family = quasibinomial, data = by_brood2)

Table 49: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

mating_duration_sec.P1 35.880 1 4.082 0.049

mating_duration_sec.P2 0.196 1 0.022 0.882

Residuals 421.956 48

Table 50: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) 1.094 0.577 1.895 0.064

mating_duration_sec.P1 -0.015 0.008 -1.913 0.062

mating_duration_sec.P2 0.000 0.003 0.149 0.882

Mating duration did not aﬀect the number of oﬀspring sired by the P2 males (F1,48 = 0.022, P= 0.882),

but P1 males with shorter mating durations sired more oﬀspring (F1,48 = 4.082, P= 0.049).

3.3.4 The eﬀect of mating duration on the number of oﬀspring sired per male genotype

glm(formula = cbind(ST_offspring, SR_offspring) ~ mating_duration_sec.ST +

mating_duration_sec.SR, family = quasibinomial, data = by_brood2)

Table 51: Analysis of variance table

Sum Sq Df F value Pr(>F)

mating_duration_sec.ST 30.422 1 3.366 0.073

mating_duration_sec.SR 2.226 1 0.246 0.622

Residuals 433.796 48

Table 52: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -1.380 0.896 -1.541 0.130

mating_duration_sec.ST 0.024 0.014 1.734 0.089

mating_duration_sec.SR -0.001 0.003 -0.488 0.628

Mating duration did not aﬀect the number of oﬀspring sired by the SR males (F1,48 = 0.246, P= 0.622).

However, it did aﬀect the number of oﬀspring sired by the ST males (F1,48 = 3.366, P= 0.073).

3.3.5 The eﬀect of mating position with mating duration on the number of oﬀspring sired

glm(formula = cbind(ST_offspring, SR_offspring) ~ mating_duration_sec.ST +

mating_duration_sec.SR + mating_position, family = quasibinomial,

data = by_brood2)

Table 53: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

mating_duration_sec.ST 32.549 1 3.565 0.065

mating_duration_sec.SR 0.535 1 0.059 0.810

mating_position 11.025 1 1.208 0.277

Residuals 429.068 47

Table 54: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -1.269 0.892 -1.423 0.161

mating_duration_sec.ST 0.024 0.013 1.797 0.079

mating_duration_sec.SR -0.001 0.003 -0.240 0.811

mating_positionP2 -0.407 0.372 -1.096 0.279

Mating position still did not aﬀect number of oﬀspring sired by each male when mating duration was included

as a covariate (F1,47 = 1.208, P= 0.277).

3.3.6 The eﬀect of male genotype with mating duration on the number of oﬀspring sired

glm(formula = cbind(P1_offspring, P2_offspring) ~ mating_duration_sec.P1 +

mating_duration_sec.P2 + male_genotype, family = quasibinomial,

data = by_brood2)

Table 55: Analysis of variance table (Type II tests)

Sum Sq Df F value Pr(>F)

mating_duration_sec.P1 35.442 1 3.950 0.053

mating_duration_sec.P2 0.052 1 0.006 0.940

male_genotype 0.636 1 0.071 0.791

Residuals 421.668 47

Table 56: Model Coeﬃcients

Estimate Std. Error t value Pr(>|t|)

(Intercept) -1.156 0.630 -1.835 0.073

mating_duration_sec.P1 0.015 0.008 1.882 0.066

mating_duration_sec.P2 0.000 0.003 -0.076 0.940

male_genotypeST 0.099 0.372 0.266 0.791

Male genotype type still does not aﬀect the number of oﬀspring sired by each male when mating during is

included as a covariate (F1,47 = 0.071, P= 0.791).

4 Supplementary ﬁgures

Figure S1: Variation in thorax length and eyespan with male genotype

Figure S2: Variation in relative eyespan with genotype

Supplementary Information 3: all model tables and eﬀect sizes for

broods with 10 or more oﬀspring genotyped