You are on page 1of 11

doi: 10.1111/j.1420-9101.2008.01616.

x

Testing multiple hypotheses for the maintenance of male
homosexual copulatory behaviour in flour beetles

K. E. LEVAN, T. Y. FEDINA & S. M. LEWIS
Department of Biology, Tufts University, Medford, MA, USA

Keywords: Abstract
dominance; Diverse animal groups exhibit homosexual interactions, yet the evolutionary
indirect sperm transfer; maintenance of such behaviours remains enigmatic as they do not directly
mating behaviour; increase reproductive success by generating progeny. Here, we use Tribolium
sperm ageing; castaneum flour beetles, which exhibit frequent male homosexual copulations,
sperm dumping. to empirically test several hypotheses for the maintenance of such behaviours:
(1) establishing social dominance; (2) practice for future heterosexual
encounters; and (3) indirect sperm translocation. We found no evidence that
Tribolium males use homosexual copulations either to establish dominance or
to practice behaviours that increase their subsequent heterosexual reproduc-
tive performance. Our results provide limited support for the hypothesis of
indirect sperm translocation: when males from two genetic strains mated with
females immediately following a homosexual copulation, females produced
progeny sired not only by the directly mating male, but also by that male’s
homosexual partner. However, this phenomenon was detected in only 7% of
homosexual pairs, and in each case such indirectly sired progeny accounted
for < 0.5% of females’ total progeny. Direct observations indicated that
mounting males often released spermatophores during homosexual copula-
tions. These observations suggest that homosexual copulations may be a
behavioural mechanism that allows males to expel older, potentially low-
quality sperm. Additional work is needed to test this new hypothesis, and to
determine whether sperm release during homosexual copulations occurs in
other groups.

Proximate genetic, neural and hormonal causes of
Introduction
homosexual behaviours have been identified for some
Homosexual behavioural interactions have been docu- model organisms (e.g. Adkins-Regan, 1998; Svetec et al.,
mented in wild populations of numerous animals, 2005; Yamamoto, 2007). However, only a few empirical
including many birds, reptiles, mammals, insects and studies have examined possible explanations for the
spiders (Bagemihl, 1999; Sommer & Vasey, 2006). For evolutionary maintenance of homosexual behaviours
example, male homosexual mountings occur in over 100 (but see Harari et al., 2000; Sommer & Vasey, 2006).
insect species (reviewed in Thornhill & Alcock, 1983; In spite of their common occurrence, the potential
Iguchi, 1996; Bagemihl, 1999; Leal et al., 1998; Serrano costs and benefits of male homosexual behaviours
et al., 2000; Switzer et al., 2004; Reinhardt et al., 2007). remain poorly understood. Homosexual behaviours are
generally assumed to be costly because they reduce
Correspondence: Dr Sara M. Lewis, Department of Biology, Tufts University, opportunities for heterosexual matings and associated
Medford, MA 02155, USA. progeny production. Several hypotheses have been
Tel.: +1 617 627 3548; fax: +1 627 627 3805; proposed for the apparent paradox of homosexual
e-mail: sara.lewis@tufts.edu
behaviours (Thornhill & Alcock, 1983; Bagemihl, 1999;
Present addresses: K. E. Levan, Division of Biological Sciences, University
of California-San Diego, La Jolla, CA 92093, USA. Harari et al., 2000; Sommer & Vasey, 2006). Male
T. Y. Fedina, Department of Ecology & Evolutionary Biology, University homosexual behaviours might provide a net fitness
of Michigan, Ann Arbor, MI 48109, USA. benefit if: (1) such behaviours help establish social

ª 2008 THE AUTHORS. J. EVOL. BIOL. 22 (2009) 60–70
60 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

She further speculated that the accretion represented flour mixed with seminal fluid deposited during homosexual copula- tions. Castro et al. females. 1982. Serrano et al. 1996. Fedina & was significantly shortened (median of 15 weeks) com- Lewis. (1994) found a response to artificial selection been proposed. 2007). 1991. but presented no evidence to support this idea. Two studies demonstrated a genetic as high sexual activity. Tribolium castaneum males also discriminate ª 2008 THE AUTHORS.. such previous studies. 1972. as they preferentially contact and mate with and hind-legs along the sides of the mounted beetle. These male–male copulations closely resemble hetero- ual copulations allow males to indirectly transfer their sexual copulations (detailed descriptions of the latter are own sperm to females through a male intermediary. although anecdotal reports of homosexual (1991) used a diallele cross among inbred lines. and speculated that this accretion might be responsible for reduced longevity by blocking the external opening of the digestive tract. 2008). In addition. Faustini & Halstead. Thus. or (5) such behaviours are iour in Tribolium castaneum have been characterized in genetically correlated with a positively selected trait. King & Dawson. Male homosexual pared with both males kept alone (median of 50 weeks) copulations (see Fig. 2000). even in the absence of a net fitness benefit Lewis. 1995).. whereas behaviours are common and various hypotheses have Castro et al. which provide an excellent mating rate. Previous researchers have suggested that Tribolium homosexual copulations represent a failure of sex recognition (Castro et al. Bloch Qazi. and their natural environment can in homosexual copulations. The latter study alternative hypotheses for their evolutionary mainte. Wool. 1981. also tested whether homosexual behaviour is maintained nance. Scale virgin females (Graur & Wool. Lewis & Iannini. easily be recreated in the laboratory (reviewed by Park. this seems unlikely based on marked chemical and morphological differences between the sexes in Tribolium. 1967.. Graur & Wool. extending its genitalia downward and forward to contact distinguishing between virgin and previously mated the posterior end of the mounted male (below). 2008). provided by Sokoloff. Fig. leading to higher reproductive success. However. Tribolium adults mate frequently throughout a lifespan of 1–2 years (Sokoloff. females. as males possess conspicuous setiferous glands located on the femurs of their prothoracic legs (Faustini et al. white accretion at the posterior end of their abdomen. Mate attraction appears to rely primarily on volatile chemical signals produced by these and other glands (reviewed by Fedina & Lewis. 2003. Serrano (2) males gain practice that enhances their subsequent et al. 2008). 1980. Homosexual copulations in flour beetles 61 dominance. and this behaviour occurs in both same-sex and mixed. and engaging in leg-rubbing behaviour. 1 Homosexual copulation between Tribolium castaneum males: mounting male (above) exhibits typical heterosexual copulatory T. 1) are frequent in Tribolium spp. Tribolium flour An earlier study by Spratt (1980) found a marked beetles are cosmopolitan stored grain pests that have reduction in adult longevity for males kept in all-male long been popular for studies of population ecology and groups (under these conditions males frequently engage sexual selection. Fedina & Alternatively. Here. few studies have systematically tested for increased homosexual behaviour. 1994). we report such studies conducted on through a positive genetic correlation with overall Tribolium flour beetles. 1967. extend- (4) males either cannot distinguish between males and ing its genitalia.. 1982. 1974. sex groups (Wool. or (3) homosex. Spratt. During heterosexual copulations. castaneum males transfer sperm to females inside a spermatophore produced by two pairs of accessory glands (Bloch Qazi et al. 1982).. J. She also noted that males kept in same-sex groups often developed a hardened. 1974. component to male homosexual behaviour: Serrano et al. 2000): in all-male groups. Sokoloff. male longevity 1948.. Serrano et al. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . or do not distinguish because the cost of lost Some important aspects of male homosexual behav- mating opportunities is low. copulation rates correlated with overall sexual activity. heterosexual reproductive performance. 1974).. Pai & Bernasconi. 1994. and rubbing its mid. EVOL. bar = 1 mm. 2000). but found no evidence that homosexual model system for studying the potential fitness conse. quences of homosexual behaviours. Tribolium male homosexual copulations homosexual behaviours might be maintained if: involve a male dorsally mounting another male. castaneum males have been shown to be capable of behaviour. as suggested for homosexual copulations in other insects (Harari et al.. or males kept in mixed-sex groups (median of 40 weeks).. Fedina. 2008. Tribolium are sexually dimorphic. BIOL. T.

Lewis & Austad. Such reproductive benefits might occur provided with excess flour. or through longer used a standardized density of one beetle per 2 g flour). Data were analysed using S P S S 16. Each male was contact the posterior abdomen of the other male. which is agonistic behaviours that might be associated with that previously isolated males should be more likely to establishing social dominance (Sokoloff. We tested one prediction of this Although Tribolium beetles exhibit no overt intermale novel explanation for homosexual behaviour. In addition. and allowed to mate with five successive virgin females. pairs). 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . Following described by Spratt (1980) are formed by male seminal established methods (Schlager. products. against mating females with whom they have previously females. we predicted that certain males would consistently mount other males. these pairs we designated as ‘mounting’ those males 2003). if rapidly rubbing its legs along the sides of the mounted males did not mate with a given female within 5 min. semidominant black body colour allele. 1996). BIOL. (SPSS Inc. insemination based on the presence or absence of larvae In this study. If males use homosexual behaviour to establish dominance.. during copulation males rapidly mounted or was mounted. Lewis. we initiate homosexual copulations compared with males hypothesized that males might establish dominance by kept in same-sex groups. on males’ repro. we measured male insemination success (pro- it seems unlikely that lack of either production or portion of copulations leading to insemination) by perception components of sexual recognition accounts isolating each mated female for 2 weeks and scoring for homosexual behaviour in these beetles. ‘mounted’ male. we hypothesized that Hypothesis 1: social dominance homosexual copulations might be a mechanism for males to discard old sperm.g. rub their legs against the sides of the female’s body one of the two males mounted more frequently. we observed male pairs during homosexual Chicago black (b ⁄ b) which is homozygous for an autoso- copulations to test the hypothesis that the accretions mal. whereas the remaining male was designated as the ing information for a video illustrating these behaviours). LEVAN ET AL. this hypothesis predicts Beetle maintenance and general methods that dominant mounting males would show higher All beetles were sexed as pupae and housed in a dark reproductive success in their subsequent heterosexual incubator at 29 C and 70% humidity. In a subset of these pairs.0. during tion. we used Tribolium casteneum to experi. previously linked to reduced longevity. 2006). that would develop only if insemination had occurred. copulations were remarkably similar (see online Support.2 to 11 min. Chicago ductive performance. then dorsally mounting another male in the correct orienta. Kravitz. Finally. so. following Lewis with other males (Arnaud & Haubruge. castaneum nance of male homosexual copulatory behaviour: social genetic strains differing in adult body colour: a wild-type dominance. copulating with other males. EVOL. based on our observations IL. extending its aedeagus downward and forward to their heterosexual reproductive success. between progeny sired by either black or wild-type males.62 K. Pairs were observed continuously for 1 h. E. Behaviours exhibited during male homosexual that mounted at least twice often as the other male. showing that mounting males do expel seminal products during homosexual copulations. (2004). body mass. e. practice and indirect sperm translocation. This subset of males was kept together these are defined throughout this paper as a male in pairs (in 1 g of flour) for an additional 23 h. number of copulations and insemination Male heterosexual copulation rate was measured by success were compared between pairs of mounting and providing males with a succession of individual virgin mounted males using paired t-tests. we used the formed differentially on the mounting or mounted males. Following previous studies (Lewis & Virgin males were marked individually with paint Iannini. term effects of dominance status (perhaps hormonally Mating observations were conducted at 29–32 C in mediated. Therefore. Fedina & Lewis. 1960. Adults were encounters. 2000) on mating and insemina- plastic arenas with scratched bottoms or filter paper to tion success. and kept either individually either immediately by enhancing dominant males’ suc- or in groups depending on the experimental design (we cess in a competitive mating situation. J. we defined heterosexual copulations as a dots and randomly assigned to male–male pairs (n = 26 male dorsally mounting a female in the correct orienta. Not all copulations result in successful insemi- mated in favour of females that had previously mated nation (Bloch Qazi et al. 1995). Male copulations ranged from 0. provide traction. and Methods that dominant males might show a size advantage over subordinate males. isolated for 24 h before weighing males and comparing tion. male’s body. Experimental beetles came from two T. extending its aedeagus and inserting it into the which we recorded the number of times each male female’s genital opening. 2004. USA) after verifying assumptions of statistical tests. mentally test three adaptive hypotheses for the mainte.. 1974). body colour phenotype of adult progeny to distinguish We further examined the effect of these accretions. ª 2008 THE AUTHORS. and to determine whether such accretions are 1990. The duration of such male homosexual she was replaced with another virgin female. In strain (+ ⁄ +) homozygous for a chestnut body allele. 1999). for interspersed with periods of inactivity (Bloch Qazi. and addition.

After mating with the focal (wild-type) male. if no copulations. nine males each). mating rate (number of matings 15 min)1) trapped under spines on the male aedeagus. domly assigned to pairs (n = 16 pairs. Only females with confirmed insemina. equally as mounting males). To determine whether homosexual copulations (Fig. castaneum individually to determine insemination success (propor. At 1–2 weeks intermediary could be facilitated by sperm becoming post-eclosion. Randomly selected. We scored female was allowed to lay eggs for 3 days (oviposition body colour phenotypes (as b ⁄ b or b ⁄ +) of all adult period 1). 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . Again. Thus. These authors pupae to be kept either isolated (n = 20 males) or in proposed that such translocation through a female groups (two groups. it is was measured by offering each male a sequence of single possible that similar mechanisms could result in indirect virgin black females. Homosexual copulations in flour beetles 63 We conducted a separate experiment to test whether group males. pairs. we used the type focal male. Immediately after each We also examined whether isolated males differed from male homosexual copulation. we compared the mating rate but distinct phenomenon was studied by Haubruge et al. both strains were represented approximately nation success compared with isolated males. To reared to adults. After the mounting male dismounted. A related examine this hypothesis. ª 2008 THE AUTHORS. we used nonparametric Mann–Whitney tests successful than mounted males when they competed at each oviposition period to compare paternity share of directly for access to females. the female was replaced We conducted an experiment using males from differ- with a new female. a single female was placed in the arena and we recorded Males might engage in homosexual behaviour to transfer which male successfully mated. during the initial homosexual copulation. We used an exact their sperm indirectly to females through a male inter- binomial test to determine if homosexually mounting mediary. mating occurred within 5 min. each with one marked male) and observed until a homosexual copula- Hypothesis 3: indirect sperm translocation tion occurred. To control for any progeny that might arise due ed with a black male for 24 h and then isolated for 24 h to back ⁄ forward mutations between the black and wild- (the presence of progeny from this intermating vial was type alleles. and male). males that had previously been males removed sperm from previously mated females. we also mated single black or wild-type used to confirm successful insemination by the first males with virgin black females (n = 65 matings). EVOL. kept in groups (thus engaging in frequent homosexual and that some of this sperm was subsequently transferred copulations). Because these data were not normally the mounting males from homosexual pairs were more distributed. who reported evidence that mating T. To homosexual copulations in any other insects. All progeny were control males (27 180 total progeny were scored). Xylocoris maculi- pennis: sperm from the mounting male apparently migrated into the seminal vesicles of the mounted male. Wild-type males were randomly assigned as when they mated with a different female. each collected eggs from each female as above. Carayon (1974) described transfer of male and mounted males were equally likely to mate when sperm through traumatic insemination during homosex- they competed for access to a single female. virgin black males and variance t-tests to examine the prediction that males kept these male pairs were allowed to copulate once (n = 86 in groups would have higher mating rates and insemi. group males with isolated males. males engaging in homosexual copulations might indi- tion of copulations leading to insemination) as described rectly inseminate females through a male intermediary above. and the insemination success of males that had been (1999). Each mated female was kept ent genetic strains to examine whether T. and second-male paternity share was detect progeny fertilized by any sperm translocated calculated as the proportion of progeny sired by the wild. virgin wild-type males were might provide a practice advantage. ual copulations in the anthocorid bug. mounted and additional 7 days (oviposition period 2). Virgin males were ran. it is unknown whether any sperm thus passed encounters to this male intermediary is later transferred to females Homosexual copulations might provide behavioural during subsequent copulations by the mounted male. These mated females were isolated a group for 1 week. wild-type males were isolated to copulate once with a virgin black female randomly assigned as pupae to be kept either isolated or in (total of 172 matings). castaneum previously isolated vs. or practice that would increase a male’s reproductive whether such indirect sperm transfer occurs during success in subsequent heterosexual encounters. males were separated and group males in their paternity share when mating with then both the mounting and mounted males were previously mated females. then transferred to a new container for an progeny from matings of the mounting. Hypothesis 2: practice for subsequent heterosexual However. presence of progeny showing unexpected body colour tion by both males were included in the data analysis: this phenotypes based on the known genotype of the directly resulted in final sample sizes of 25 isolated males and 17 mating males. we used separate paired with randomly selected. 2 weeks. J. females were sperm translocation by males engaging in homosexual removed and replaced with a new female. BIOL. also. 2). After each mating. Each focal male was allowed to to oviposit with weekly transfers to new containers for copulate once with a black female that had been premat.

Fig. 2 Experimental design used to test the indirect sperm translo- Because neither insemination rate nor insemination cation hypothesis. For the 12 pairs that engaged in vations we measured the number and duration of homosexual copulations. same-sex groups of 20–30 males or isolated males wild-type males (n = 14 pairs) were introduced into a (beetles were marked with small paint dots to distinguish short length of glass tubing 4 mm wide.25 mm2) and without accretions: each male was placed with five wild-type virgin females. We used A N O V A to examine differences among these three male groups in their total number of contacts and copulations 10 min)1. and thus decrease In 32. no reproductive advantage was evident when ª 2008 THE AUTHORS. After homosexual copulations between virgin males from two genetic strains (n = 86 pairs). sperm (Reinhardt. To obtain success (Fig. paired t = )0. mounting likely to show accretions. J. We hypothesized that mounting males might cause accretions to form on. paired heterosexual reproductive performance. Mating behaviour was measured over 10 min for males with accretions (minimum size 0. To examine this hypothesis. we observations. so. One prediction of this novel hypothesis is that males kept isolated should be more likely to initiate Accretion formation and effects on reproductive homosexual copulations than males kept in groups. BIOL. LEVAN ET AL. Indirect sperm males do release seminal products during homosexual translocation was detected as the presence of progeny indirectly sired copulations. performance presumably because the former lacked any previous We observed male–male pairs to determine how the opportunities to discard sperm. 0.2 ± used an exact binomial test to evaluate the null hypoth.396. 18 d. 3b: paired t = 1. tions might be a mechanism that allows males to get rid of old. EVOL. wild-type males together in groups with flour for 3–5 months. yet it is not known whether the accretion is formed on the mounting male.64 K.f. 18 d. 2007). however. To test this prediction. In the remaining pairs.f. there was no significant difference each) were placed individually in 2 g of flour. P = 0. other males thereby incapacitating their roles between mounting and being mounted during 1 h reproductive rivals.180). 1980) are formed. P = 0. such that no clear designation could be observed 16 male pairs until copulation occurred. Pairs of virgin. Each time a mating occurred. when tested following 24 h in isolation. we hypothesized that homosexual copula- by the mating male’s homosexual partner.. used beetles that had been kept for 3–5 months either in male groups (Spratt. males with and without accretions. We observed pairs (n = 24) under 30· magnification until homosexual copulations interacting in 1. Each male’s insemination rate (number of inseminations 10 min)1) as well as insemination success was measured as described above. and during 10 min obser- occurred or 1 h elapsed. E. the mounted Hypothesis 1: social dominance male or both. these males were assigned to a third treatment group.487). and mounted males did not differ in either the number of We also examined the effect of male accretions on heterosexual copulations they achieved (Fig. one male mounted more which both the mounting and mounted males (n = 16 of frequently.5-cm2 arenas. Males were in body mass between these mounting males (mean ± SE checked for accretion formation after 15 days. Females were isolated to oviposit for 2 weeks and adult Because our observations indicated that mounting progeny phenotypes were scored after 45 days. mounting males (here success met normality assumptions. Tribolium accretions begin to appear in all-male groups Results after several weeks. as well as in the number of copulations per contact (data were checked for homoscedascity and normality). we noted whether and where homosexual mountings by each male..05 mg) and the males they mounted (2. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . specifically male t = 1. we used nonpara- shown as + ⁄ + for illustrative purposes only) and mounted (shown metric Kruskal–Wallis tests to compare these aspects of here as b ⁄ b) males were each allowed to copulate with a virgin b ⁄ b reproductive performance among male groups. some males’ accretions fell off.08 mg.. the mounting male deposited any material. the mated female was removed and replaced with a new virgin female. and we 2. female. and observed these two treatments).710. after made. and the number of male-initiated contacts and copulations was recorded.163) or their insemination mating behaviour and insemination success.1% of the 26 male pairs.f.455. we kept 30 virgin Similarly. More- esis that mounting and mounted males were equally over. 3a. During these mating observations. males frequently switched longevity of. 17 d. P = 0. we accretions previously shown to be associated with all.1 ± 0. and potentially damaged.

0 0 0. 36 d. female.6 3 2 0. group males had significantly lower copulation rates (Fig.5% of each females’ total progeny. 4 Heterosexual reproductive performance measured immedi- Male position Male position ately after Tribolium castaneum males were kept in isolation (open bars. oviposition days 0.50 1–3: Mann–Whitney U = 175. ª 2008 THE AUTHORS.623. females produced progeny with body colour phenotypes indicating that they were sired such males were tested in mate competition for a single indirectly by their mate’s previous homosexual partner. each to mate with a virgin female. measured as (a) the Fig. occurred during the initial homosexual copulation. n = 18).25 0.6 # Copulations 0.0 Isolated Group Isolated Group d g d g Male treatment Male treatment u nte unti n u nte unti n Mo Mo Mo Mo Fig.0 5 4 0. In six of 86 pairs (7%). EVOL.5. Because P2 changes over time. oviposition days 4–10: Mann–Whitney U = 191.f.665) 0. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY .2 0 0.4 2 1 0.402).. Homosexual copulations in flour beetles 65 (a) (b) (a) (b) 5 1. measured 24 h after homosexual encounters as (a) the number of heterosexual copulations (of five females offered) and (b) insemi- nation success (proportion of copulations that resulted in insemi- nation).4 0. cases the indirect sire was the mounted male.2 1 0. Mean + SE are shown. In each case only a single progeny was indirectly sired.8 Insemination success Insemination success 4 # Copulations 3 0. colour phenotypes of all adult progeny resulting from data are presented separately for oviposition period 1 (1–3 days after these subsequent heterosexual matings. However. P = 0. None of Hypothesis 2: practice for subsequent heterosexual the 65 control matings produced any unexpected prog- encounters eny (of 11 765 total progeny scored).75 or in their last-male paternity share when they mated with previously mated females (Fig. contrary to the predictions of the practice hypothesis. indicating a low When males that had been kept in all-male groups (thus probability that these represent back mutations from engaging in homosexual copulations) were compared black to wild-type or vice versa. 3 Heterosexual reproductive performance of Tribolium castane- number of heterosexual copulations per 15 min and (b) insemina- um males as a function of male position during prior homosexual tion success (proportion of copulations resulting in successful encounters. 4b: separate variance t = 0. with same-aged males that had been isolated. 0.0 1. repre- senting 0. P = 0. we could detect the second mating) and 2 (days 4–10).590). whereas in the remaining three (n = 16 pairs. J. 4a: separate variance t = 2. BIOL.013). n = 20) or in groups (shaded bars. P = 0..f. mounting (nine males) and mounted males In three of these cases the indirect sire was the homo- (seven) were equally likely to copulate with females sexually mounting male. exact binomial test P = 0. 5 Boxplots for paternity share (P2 = proportion of progeny sired lation between males from two different genetic strains. P = 0.8 0.00 males did not differ in their insemination success Paternity share (Fig. Each male’s heterosexual reproductive performance was insemination). 1. By inspecting body n = 25) for 1 week prior to testing.00 Hypothesis 3: indirect sperm translocation 1 2 To detect indirect sperm translocation during homosex. n = 17) or isolated (open boxes. immediately after a homosexual copu- Fig.336. by focal second male) of Tribolium castaneum males that were either we allowed the mounting males and the mounted males kept in groups (shaded boxes.437. 36 d. Oviposition period ual copulations. Mean + SE are shown for n = 19 mounted males (open whether sperm translocation between males had bars) and n = 19 mounting males (solid bars). 5.

87. P = 0. male insemination success (proportion of copulations that led to mounting males released spermatophores (Fig. of contacts 12. 6 Spermatophore release during male homosexual copulations in Tribolium casta- neum flour beetles: (a) mounting male is releasing a spermatophore from its extended genitalia. ª 2008 THE AUTHORS.3 (±0.0 (±1. J. likelihood for mounting males to develop such accre. number of copula- In 8 out of 12 male–male pairs (67%) observed under tions and number of inseminations during 10-min observations and 30· magnification during homosexual copulations. LEVAN ET AL.188. cantly longer durations (mean difference 5. Mating behaviour and insemi.1) 0.2 ± 0. isolated sexual reproductive performance.5 (±0.f. and remained mounted for signifi- initiated contacts (A N O V A F2.4 (±0. tions.2% of all same-sex groups until many of them had developed observed homosexual copulations (80 of 84 homosexual accretions (3–5 months). P < 0.063).5 mm.2) 11.3.599.23 = 0. 23 d. BIOL. the Accretion No accretion Accretion dislodged mounting male or some other surface. Males in these difference 3.558).23 = 1. P < 0. groups did not differ in their mean number of male. of copulations 3. Isolated males without accretions and males whose accretions got males also had significantly higher mounting rates (mean dislodged during the experiment (Table 1).6 (±0.0001). few studies have addressed evolutionary and this was significantly reduced relative to males whose aspects of this phenomenon.011). the sticky surface of No.4) 2. P = 0.598. of inseminations 0. in homosexual copulations). males were kept in males were the mounting individual in 95. P = 0. flour beetles. we conducted a series of experiments that ferroni t = 2.0001) compared with males insemination success (Kruskal–Wallis H = 4. In this study on Tribolium accretions were dislodged during the experiment (Bon.23 = 3.5 (z = 8. and then accretions and five males whose accretions were dislodged during moved each mounting male and mounted male the experiment are shown.5 min. To determine Insemination success 0.3) 9. previously kept in groups.1) flour and develop into the accretions that form on males No. copulations). (b) a spermatophore deposited on a mounted male after release by the mounting male..103.. 11 males without males to engage in homosexual copulations. males with accretions achieved significantly different than expected by chance (exact binomial test.f.66 K. Although not significantly P = 0.613.3) these spermatophores is likely to become coated with No. substantial flour- covered accretions were found on six mounting males and on one mounted male.3 (±0.01). this indicated a trend toward a greater whose accretions were dislodged (Dunn’s test. After 15 days. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . 23 d. scale bars are 0. mea- sured as the number of male-initiated contacts.100). 6) that successful insemination). When homosex- ual copulations take place in flour. Arrows indicate spermatophores. paired t = 6. ber of copulations (A N O V A F2.24.693.2) kept in same-sex groups (Spratt. Accretion formation and effects on heterosexual Table 1 Effects of accretions resulting from Tribolium castaneum reproductive performance male homosexual copulations on male reproductive fitness.6 (±4.3 (±0. male Discussion groups differed significantly (Fig. apparently paradoxical behaviour.1) 0. fewer inseminations 10 min)1 compared with males P = 0.041). 1980).6 (±1.2) which males developed accretions.323) or paired t = 9. individually into flour. E. many animals.0001). (a) (b) Fig. P = 0.3 (±0.5) 4. In such pairs. num.6) 4.8 (±2. which was significantly greater than a nation success was compared for males with accretions. P < 0. when the proportion of contacts that led to successful copulation was compared. P = 0.016). opposite of what would be predicted if males they were much more likely to initiate homosexual initiated homosexual copulations to incapacitate their copulations when paired with males that previously had reproductive rivals.4 (±1.2) 1. P < 0. EVOL. However. males with accretions showed reduced ability Although homosexual behaviours have been described in to successfully copulate with the females they contacted. became attached to either the mounted male. We found that after males had been kept isolated.1 ± 0. the number of allowed us to simultaneously examine multiple inseminations achieved by males was significantly differ. hypotheses that have been proposed to explain such ent between groups (Table 1: Kruskal–Wallis H = 9. been kept in same-sex groups (and thus had engaged To examine the effects of accretions on males’ hetero. 7: A N O V A F2. In addition.5 mounts per 10 min. we allowed 16 pairs of Mean ± SE based on 10 males with accretions. random expectation of 0.

or in their paternity share when they copulated 0. BIOL. Horizontal lines represent mean values and letters indicate subgroups that differ significantly by Bonferonni multiple comparisons. attributed male homosexual copulations to mistakes in How might such indirect sperm translocation happen? sex recognition. male lacked such accretions (n = 11). castaneum copulations. 0. as + ⁄ b het- mounted males tested in either competitive or erozygotes are easily distinguished from b ⁄ b homozy- noncompetitive mating situations. the indirect sires were equally likely to be mounted frequently switched roles. have provided progeny whose body colour phenotype suggested they evidence that homosexual behaviours can alter social were sired not by the directly mating male. other social functions T. These unexpected progeny are reproductive performance between mounting and unlikely to represent genotyping mistakes. in the horned trapped under penile spines might be implicated in such beetle Allomyrina dichotoma. In addition. male (rather than through a female) intermediary. but rather by interactions by facilitating alliances or by establishing that male’s homosexual partner: however. Although our study focused on the male homosexual mountings may inhibit intrasexual distinct phenomenon of sperm translocation through a aggression. Based on the lack of courtship or Haubruge et al.0 Accretion None Dislodged with previously mated females. females that mated with males imme- greylag geese. it is plausible that males kept in groups might successfully copulate with females they contacted during 10-min alter their behaviour in response to higher perceived observation periods for three types of males that had been kept in sperm competition risk. such an effect seems same-sex groups for 3–5 months prior to the experiment. a ª 2008 THE AUTHORS.g. Indirect sperm translocation hypothesis Our results provide limited support for the hypothesis of indirect sperm translocation. When 0. including flamingos. rather than decrease. Also counter to the gotes. 2001). only a single dominance relationships. unexpected progeny are unlikely to dominance hypothesis. Studies on several vertebrates. In six instances. Males unlikely because increased sperm competition risk would either had hardened accretions on their posterior body (n = 10). which proposes that males engaging in homosexual copulations might indirectly Social facilitation and/or dominance hypothesis transfer sperm to females through a male intermediary.0 Practice for future heterosexual encounters a. we found no unexpected progeny was observed in each case. although.6 a compared with same-age isolated males. Iguchi (1996) suggested that indirect fertilization. EVOL. In addi- copulations to establish dominance.8 sexual copulations provide T. as females are generally larger. the authors (Futuyma. dislodged during the course of the experiment (n = 5). repre- evidence that Tribolium males use homosexual senting 0. alliance formation and conflict between sequential females.4 with females. social bonding. However. as paired males tion. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . be expected to increase. J. as the observed rates are small males typically mounted larger males during substantially above the typical mutation rates for phe- homosexual copulations in the weevil Diaprepes notypic mutations of 10)4–10)6 per locus per generation abbreviatus. In fact. males kept in groups had signifi- cantly lower mating rates (number of heterosexual copulations per 15 min) compared with isolated males. (1999) reported that in heterosexual agonistic interactions in Tribolium. (2000) found that have arisen from mutations. males kept in same-sex groups (thus engaging in homosexual copula- tions) did not show better reproductive performance 0. other factors may alter the behav- iour of group males compared with isolated males. and suggested that sperm avoidance) also seem unlikely. and for those pairs with more or mounting males. 7 Differences in the ability of Tribolium castaneum males to example.b b hypothesis We found no evidence to support the idea that homo- 0. 2006). Homosexual copulations in flour beetles 67 1.5% of each females’ total progeny. For Fig. or had accretions that were mating rates (Simmons. In addition to increased Male groups homosexual activity. suggesting the possibility of bidirec- consistent roles we found no differences in body size or tional sperm transfer. 2005). Harari et al. rhesus macaques diately following a homosexual copulation produced and bonobos (Sommer & Vasey.2 There were no differences between groups in their insemination success once they copulated with virgin females. sperm may be translocated (e. However. castaneum males with behavioural practice that increases their reproductive Copulations/Contact success in subsequent heterosexual encounters. bottlenose dolphins.

this idea will require much further testing. able (e. although our results are not fuelled by adult nutrient input and adults mate repeat- conclusive. production rates provide a selective advantage. masturbation and spontaneous sperm emission research assistance. Observations of spermatophore interactions with females. EVOL. high sperm quently mates with a female. (2000) found result in selection for continuous sperm production. Although Spratt (1980) found reduced longevity for In summary. Harari et al. Our direct observations of esis is consistent with two observations in the present homosexual copulations under high magnification sup. Thus. Such a reproductive advantage for young sperm could select not only for sperm dumping associ. Our results evidence that homosexual copulations are used to indicate that these accretions do physically interfere with establish dominance. Although of course males when they are kept in same-sex groups (Sokoloff. However. and we are grateful to them for their expert 2005). Additional work is needed Possible advantages of sperm dumping to experimentally test this hypothesis. and Rein. males with accretions had diffi. J. 2004). potentially low-quality sperm. We found that both isolated males were much more likely to mount other mounting and mounted males developed flour-covered males and exhibited higher rates and durations of accretions by 2 weeks following a single homosexual homosexual copulations relative to males from in copulation. castaneum males with a develop on the posterior abdomen of some T. 1980. and numerous teran X. these males may be physiologically incapable performance of down-regulating their sperm production when lower This study provides additional insight into the causes and population densities are encountered. hardt & Siva-Jothy (2005) found that younger sperm are more likely to be stored by female house crickets. Spratt. 2008). yet such interactions have been accretions themselves might be responsible. fertilizing ability and sperm in understanding the apparent paradox of animal homo- competitive ability) declines with sperm age. Future ural mechanism that allows males to discard sperm. perhaps by poorly studied. castaneum beetles we found no blocking the anus. similar mechanism might be responsible. Thus. Sperm dumping may also be selected when during homosexual copulations. homosexual consequences of the hardened. 2003. Julia Blum. Acheta Acknowledgments domesticus. edly during their 1. When many females are avail- sperm that get transferred when either male subse.g. We also thank Natasha Tigreros. We found limited support for resulted in their successfully inseminating fewer females the hypothesis of indirect sperm translocation to females relative to other males. including Greg Tully. that increases males’ reproductive success in subsequent culty copulating with females they contacted. castaneum males can mate with up to seven different females within 15 min (Lewis. or to provide behavioural practice males’ mating abilities. indeed. Many dedicated Tufts undergraduates contributed to the ated with homosexual copulations. but also for auton. in human males (Baker & Bellis. In T. ª 2008 THE AUTHORS. study: (1) males tested immediately after removal from port Spratt’s (1980) suggestion that these accretions form same-sex groups had lower heterosexual copulation when spermatophores released by the mounting male rates compared with isolated males. research approaches that rely on empirically testing Reinhardt (2007) reviewed evidence that sperm perfor. homosexual behaviours commonly occur males kept in same-sex groups. was merely correlative.68 K. Accretion formation and effects on reproductive however. T. which heterosexual encounters. this study). same-sex groups. Kaplan. BIOL. and this phenomenon male accretions were often dislodged during males’ deserves further study. Michelle Schenk and David (Kumashiro et al. research effort described here. and (2) previously become coated with flour over time. multiple hypotheses should lead to more rapid progress mance (including motility. and so any reproductive or release during homosexual copulations led us to propose survival handicaps they entail may be temporary. castaneum mechanism for discarding sperm. and to determine These observations suggest a novel hypothesis for the whether seminal products might be released during male maintenance of homosexual copulations as a behavio. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . maculipennis. omous spermatophore expulsion in male crickets Ryan Sato. this hypoth- 1974. Reinhardt & Siva-Jothy. although it is not males cannot down-regulate their sperm production known whether any such sperm is later transferred to rates. Using irradiated males. 1993). Carayon (1974) described male-to. abbreviatus. LEVAN ET AL. E.. On the flip side. In the hemip.to 2-year adult lifespan (reviewed by lations Tribolium males exchange small quantities of Fedina & Lewis. other sexual behaviours as suggested by Reinhardt male sperm transfer through traumatic insemination (2007). as no support for indirect sperm translocation in the weevil they are income breeders whose gamete production is D. that homosexual copulations may represent a behavio- ural mechanism that allows males to discard older. it is possible that during homosexual copu. Tribolium beetles have life-history features likely to females. we also found that through a male intermediary. under high population densities). homosexual interactions in other species. white accretions that copulations may provide T. sexual behaviours. the evidence that the in diverse animal groups.

Rev. 1960. ejaculate adjustment by males and the function of masturba. Ecol.K. Ecol.C. Am. C.M. Behav. Competition between populations of the flour beetles. 2008. Herbeck. 1993.C. 18: tera: Tenebrionidae). Am. The genetic sperm in the house cricket Acheta domesticus..L. Tsuji. 24: 1817–1829. Am. Evolution.. castaneum: artificial selection. Chem. Dobzhan. Behav. Tribolium castaneum: Diallel Analysis. Mechanisms I. Lewis. B. V. variation in sperm precedence in red flour beetles. King. M. K. & Iannini. A. M.. Ecol. 2000. 135: 351–359.H. Intrasexual An Evolutionary Perspective. New York. S. Dynamics and genetics of mating Press. beetles Tribolium castaneum and Tribolium confusum. Bloch Qazi. Sociobiol. 82: 375–393. and their interaction on the courtship behaviour dichotoma septentrionalis (Coleoptera: Scarabaeidae). Behav. A potential mechanism for cryptic B Mol. Fitness consequences of differences and Natural Diversity. Proximal traits and mechanisms Serrano. 26: 361–367. Serrano. Higuchi. 1990. 2008. R.. J. for discussions and comments that substan. Spratt. Bagemihl.R.N.. A. Biol. 17: 1–16. J. C. M. 2004. Population biology and the Effects of environmental and social conditions on homosexual Tribolium model. 1999. M. Cam- mounting in the beetle Diaprepes abbreviata. Sci. & Halstead. B. D. K. 2003.J. 1982. Sperm precedence in the fertilization of eggs Tribolium castaneum: site of aggregation pheromone produc. 85: 183–193.T. Behav. R. pp. Proc. R. Forsythe. S. & Lo´pez-Fanjul. Monogr.E. social Iguchi. & Vasey. BIOL. Marino. Evolutionary consequences of sperm cell homosexuelle chez Xylocoris maculipennis (Hem. Mating behaviour and male Lewis. 1948. 67: 799–804. Polyandry and female control: the tion. Sci. J. St Martin’s Press. tially improved the manuscript. S. Q. organ of a traumatically inseminating insect. B bridge.L. Hormonal mechanisms of mate choice.M. Reinhardt. J. J. Res. Biol. T. & Bernasconi. Soc. Exp. Ono. red flour beetle Tribolium castaneum as a case study. Nat. M. 21: 547– J. 1995.M. Stored Prod. L.B. J. J. W. & Meinwald. D. The Evolution of Insect Mating Kravitz. M. Res. Proc. Homosexual copulations in flour beetles 69 Philip T. Am. MA. Ecol.A & Lo´pez-Fanjul. NJ. J. S. Sinauer Associates. Leal. in Tribolium castaneum. J. Dev.and intraspecfic sexual discrimination in the flour castaneum. 558. eds). Zool. Evol. S. Antho.R. L. Arnaud.L. Evol. Biological Exuberance: Animal Homosexuality Lewis. 310: 148–159. Sunderland. 266: 1183–1187. Klaus Reinhardt and an anonymous amine neurons in a complex behavior. N. & Dawson. Reinhardt. Nat. 206: 4507–4519. Am. Escajeda. EVOL. M. Effect of genes.S. P. Anim. 1998. 2006. Pai. Park..L.. Jpn. Tenebrioni. 221–238.Y. K. Oxford. behavior in Tribolium castaneum. Insect Behav. Multiple mating and repeated copulations: Adkins-Regan. Harvard University Press. Nat. aging. Princeton.. D.C. 1998. Sokoloff. Evol. Entomol. A. 1996. Sel. & Austad. Serotonin and aggression: insights gained Systems. K. Genet. T. Male homosexual behaviour and other factors Haubruge. 186: reviewer. Acad.M. 265–308. 5 (T. This work was funded by Kumashiro. Zool. J. influencing adult longevity in Tribolium castaneum and Fertilization by proxy: rival sperm removal and translocation T. Bloch Qazi. J. 7: 465–480. pairing in the Japanese beetle (Popillia japonica Newman). & Sakai. of transgenic Drosophila males. E. Steere. & Lewis.T. 16: 170–176. Sexual behavior of the horned beetle. Am. J. Homosexual Behaviour in Animals: Harari. 2005. J. C.. J. 89: 892–897. in male mating behaviour in relation to female reproductive Baker. & Bellis. 173: 43–72. Male-released sex pheromone of the References stink bug Piezodorus hybneri. S.J. Anim.A. Cambridge. & Siva-Jothy. & Ferveur.A. J. & Gage. Switzer. Morphol. sequences in the Insects. 1991. effects on male reproductive success in red flour beetles... Allomyrina experience. Brockman. Svetec. in a beetle. J. Princeton University Press. 60: 844–853. Starks. P. Oxford University Graur. Auto-spermatophore a Tufts University Faculty Research Award to S.J. Entomol. S.Y. Rev. Genet.M. Castro. X. 265: 2071–2079. G. Century-Crofts. selection in Tribolium flour beetles. dae).. Faustini.J. Soc.. E. 133–227. E. 1981..M. 53: 93–98. M. 1974. Behav.M. Insect Physiol. Vol. W. Sexually 931–935. J.. 2005. T. Shi. dimorphic setiferous sex patch in the male red flour beetle Schlager. Mignon. P. 53: 557–560.G. 2000. P.J. D. C. R. Burkholder. 1983.A. E. L. 1980. & Laub. Soc. Carayon. 1982. Experimental studies of inter-specific competition. Hecht & W. D. 2007.. Behaviour 136: 67–77. 2007. S. E. 2001. for biasing paternity in the red flour beetle Tribolium Inter. J.E. R. 2005. Comp. from a lobster model system and speculations on the role of ª 2008 THE AUTHORS. M.M. Insemination traumatique heterosexuelle et Reinhardt. Lond. Sources of intraspecific mate choice in Tribolium castaneum (Coleoptera. & Lewis. M. 64: 870–875. Female-limited polymorphism in the copulatory male Coleoptera. 165: properties of homosexual copulation behavior in Tribolium 718–723. & Lo´pez-Fanjul. & Alcock. G. confusum. Castro. 2006. C. sky. D.V. 1996. Ann. Genet. Thornhill. Heredity.T. 1974. In: Evolutionary Biology. Y. coridae). M. Simmons. The Biology of Tribolium. S. J. L.. Fedina. Cambridge University Press. J. An advantage for young Castro. 83: 151–171. & Krusel. female choice in a flour beetle. Cryptic female choice during spermatophore The genetic properties of homosexual copulation behavior in transfer in Tribolium castaneum (Coleoptera: Tenebrionidae). A. Harney. Ann. Fedina. 12: 161–179. 38: 166–178.. R..Y. Setiferous structures of M. Appleton. Anim.. Soc. 2003.M.C. 1999. E. Physiol. Arnaud. 170: Faustini. R. 16: 109–114. 2004. Entomol.A. & Landolt. M.. K.S. Toro. tion. Y. 46: 861–885. & Wool. extrusion in male crickets. L. Genet. 50: 1157–1160. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . Fedina. H.W. of sperm transfer and storage in the red flour beetle (Coleop.. Toro. 1999. 2000. Biol. Biol. 1972. T. C. Toro. L. 2007. Chem. D 278: 2803–2806. Houot. B Biol. Human sperm competition: status in flour beetles. H. New York. & Lewis. Gorb. Tribolium confusum and Tribolium castaneum. & Haubruge. Naylor. & Siva-Jothy.. Sperm Competition and its Evolutionary Con- Futuyma. Sommer.G. P. An integrative view of sexual 85: 142–146. Exp. 1994. Behav. Kuwahara.

59: 39–66. heterosexual copulation for behavioural comparison. Adv. LEVAN ET AL. D. BIOL. Please note: Wiley-Blackwell are not responsible for Received 29 May 2008. Genet. The neural and genetic substrates of sexual homosexual copulations in red flour beetles. Inform. Trib. 2007. 10: 182–186. Wool. J. 22 (2009) 60–70 JOURNAL COMPILATION ª 2008 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY . EVOL. accepted 25 July 2008 the content or functionality of any supporting informa- tion supplied by the authors. Any queries (other than Supporting information missing material) should be directed to the corresponding author for the article. E.70 K. Bull. and shows a behavior in Drosophila. Some observations on mating frequency in Appendix S1 Tribolium heterosexual and homosexual Tribolium castaneum strains. copulatory behaviours: this QuickTime movie illustates Yamamoto. Additional supporting information may be found in the online version of this article: ª 2008 THE AUTHORS. D. 1967.