Mating status, independent of size, influences lethal fighting in an entomopathogenic nematode

Males of many species compete over access to females by physical contests. Previous experience of mating and the opportunity to mate with a female may in ﬂ uence the motivation of males to engage in contests and the outcome of such contests. Often, prior mating results in increased aggression and probability of success. This is mediated by the effects on the male's subjective evaluation of both the resource value (RV) and his own ability to acquire the resource (resource-holding potential, RHP). Moreover, having mated may also affect a male's actual ﬁ ghting ability. In Steinernema nematodes, mated males paired with naïve males are more likely to win contests. Here we show that this advantage in mated males of Steinernema carpocapsae cannot be explained solely by the physical changes brought about by prior female contact, since males exposed to female pheromone alone developed sperm and increased in size like mated males but did not have the same advantage as them in contests with naïve males. Effects of mating on other components of RHP, such as skill or motivation, or on RV may explain the greater probability of mated males winning. We also show that mating had differential effects on the probability of a male initiating attack, depending on whether the opponent was a mated male (and thus ready to mate again) or naïve, consistent with the relative threat of the opponent as a competitor for copulations. © 2023 The Author(s). Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. This is an open access article under the CC BY license (http://creativecommons.org/licenses/

Individuals of many animal taxa compete for resources such as food or mates through contest behaviour of various kinds, sometimes resulting in injury or even death (Alexander, 1961;Enquist & Leimar, 1987;Kravitz & Huber, 2003;Moore et al., 2008). Who wins a contest is largely dependent on the relative ability of opponents to acquire the contested resource (resource-holding potential, RHP), although the value of the contested resource to the opponents affects their motivation and hence also the outcome of the contest (Arnott & Elwood, 2008;Enquist & Leimar, 1987;Parker, 1974;Vieira & Peixoto, 2013). Body size and strength, weaponry, skill and physiological condition all contribute to an animal's RHP (Arnott & Elwood, 2009;Briffa & Lane, 2017;Palaoro & Peixoto, 2022), but size is frequently the major factor, with larger size correlating with higher RHP in many animals (Archer, 1988;Brown et al., 2006;Hack et al., 1997;Hsu et al., 2006;Hughes, 1996;Morand, 2000;Neat et al., 1998;Petersen & Hardy, 1996). Contest behaviour can be affected by contestants' assessment of their own RHP (self-assessment) and of their opponents' RHP in comparison (mutual assessment; Arnott & Elwood, 2009;Chapin et al., 2019).
Prior experience, including social experience, can influence the outcome of a contest (Beacham & Newman, 1987;Jacques et al., 1996;Kimberly & Rowland, 2000). Among the best documented effects of social experience are the 'winner effect', whereby winning a contest increases an animal's probability of winning a subsequent contest by altering its self-assessment of RHP (Rutte et al., 2006;Hsu et al., 2006), and the finding that residents also generally win contests, explained by their greater familiarity with the contested resource (Kokko et al., 2006;Kapranas et al., 2020). Many other social experiences, including those of early life, may also affect the outcome of contests (Colella et al., 2019;Favati et al., 2021;Lee et al., 2014).
For males competing for mating opportunities, prior mating or other social experience of females can affect subsequent contest behaviour (Baxter & Dukas, 2017;Innocent et al., 2011;Yasuda et al., 2015;Yuan et al., 2014;Zhang et al., 2019). There are sound theoretical arguments to predict either an increase or a decrease in aggression following mating (Baxter & Dukas, 2017;Judge et al., 2010). On the one hand, unmated males may be more motivated to engage in fighting because future mating opportunities are of greater value to them, while mated males may be less willing to engage in costly fighting (Kemp, 2006). On the other hand, having mated may change a male's evaluation of the contested resource increasing his willingness to fight for a more highly valued prize (Brown et al., 2006(Brown et al., , 2007Innocent et al., 2011;Kemp, 2006). Mating may also increase the male's self-assessment of his own RHP (Yasuda et al., 2015), in a manner similar to the 'winner' effect whereby males that win a contest become more aggressive and likely to fight (Hsu et al., 2006). In empirical studies across a range of invertebrates and vertebrates, mating and/or contact with a female tends to increase rather than decrease a male's aggression and success in contests (Bergman et al., 2010;Guevara-Fiore et al., 2012;Judge et al., 2010;Killian & Allen, 2008;Kralj-Fi ser et al., 2011;Yasuda et al., 2015;Zhang et al., 2019), while for house crickets, Acheta domesticus, males with restricted prior access to females were more aggressive (Brown et al., 2006(Brown et al., , 2007. In studies where mating or sexual experience had no effect on aggression, this may be due to the counterbalancing effects on the males' motivation; future expectations of reproductive success versus subjective assessments of RV or RHP (Baxter & Dukas, 2017;Innocent et al., 2011). Since mating, even for males, may incur costs in terms of energy (Cargnelutti et al., 2022;Perry & Tse, 2013;Scharf et al., 2013), it may also alter the outcome of aggressive encounters by reducing RHP. For example, while male field crickets, Gryllus pennsylvanicus, were more aggressive after experiencing a female (whether they mated with her or not), those that copulated were less likely to win a contest than those that did not (Judge et al., 2010). In the nematode Steinernema longicaudum (a species with lethal male combat; see 'Background biology' below) mated males were more successful in terms of survival and killing than naïve males when the two were paired together in controlled fights (Kapranas et al., 2020). Interpreting these findings is complicated by the fact that mating involves both exposure to female pheromone and the act of copulation. In Steinernema, female pheromone induces physiological changes in males, stimulating sexual maturation and sperm production (Ebssa et al., 2008;Hartley, 2017). In other species, aggression of males may be influenced by gonadal development (Neat et al., 1998) or by exposure to female pheromone (Killian & Allen, 2008). In Steinernema, the gonads are large relative to somatic body size (Ebssa et al., 2008), and therefore their maturation might be accompanied by an increase in overall body size that could contribute to RHP.
Here we tested whether the physiological change induced by exposure to female pheromone alone, without the opportunity to mate, in male steinernematids could explain their higher RHP when they engage other competitors in lethal fights. We first assessed whether male sexual maturation induced by pheromone exposure is associated with an increase in their body size. We then experimentally distinguished the effects of female pheromones and the act of mating on contest (fighting) behaviour and outcome.

Background Biology
In Steinernema spp., a free-living infective juvenile seeks out and invades an insect. Once inside the insect, infective juveniles release symbiotic bacteria which assist in rapidly killing the host and digesting its tissues. The nematodes feed on the bacteria and digested host tissue and develop to adult. A single dead host may support two to three generations and result in tens of thousands of infective juvenile progeny. In most Steinernema species, infective juveniles develop into either males or females. Although the sex ratio is slightly female biased (Alsaiyah et al., 2009), females die shortly after mating as juveniles hatch inside their mother (endotokia matricida; Baliadi et al., 2004), resulting in a male-biased operational sex ratio. Males of some Steinernema engage in lethal fighting, wrapping around and compressing their opponent, which can cause paralysis or death (Zenner et al., 2014). Lethal fighting has evolved in Steinernema spp. in response to the limited mating opportunities and to secure the valuable resource of a host for the developing offspring (Innocent et al., 2011;Kapranas et al., 2016). The incidence of fighting is much higher in males that develop from infective juveniles and colonize a host in small numbers than in males of subsequent generations which typically occur in crowded conditions (Zenner et al., 2014). In at least some Steinernema species, laboratory studies have shown that males that develop in isolation have underdeveloped testes and the seminal vesicle contains no sperm (Ebssa et al., 2008). The state of reproductive development can be clearly discerned by microscopic examination through the nematode's transparent cuticle. Gonadal development of solitarily reared males can be stimulated by exposure to female pheromone (Ebssa et al., 2008;Hartley, 2017). Of the close to 100 species of Steinernema described, Steinernema carpocapsae is the best studied and is also widely commercialized as a biological control agent (Koppenh€ ofer et al., 2020;Lacey et al., 2015;Murfin et al., 2012).

Nematode Culture
Steinernema carpocapsae (All strain) cultures were routinely maintained using standard procedures by passage through lateinstar Galleria mellonella (wax moth) larvae (Kaya & Stock, 1997) at 9 C. Infective juveniles were stored in tap water at 20 C. Adults for experiments were obtained by placing surface-sterilized (0.1 % hyamine) infective juveniles individually in 20 ml hanging drops of haemolymph from G. mellonella larvae, which provides a suitable environment for development to adulthood (Kaya & Stock, 1997;Zenner et al., 2014). The hanging drops were placed on the under surface of a petri dish lid which was placed over a water-filled petri dish (3.5 cm diameter) to prevent desiccation. The infective juveniles in the hanging drops developed to adulthood within 3 days at 20 C. Adult males can be distinguished by their copulatory spicules and their smaller size whereas females are larger and bear a vulva. We used 3-day-old adult males and females in our experiments. At this time a minority of males had spontaneously developed sperm and were discarded.

Conditioning
Males without sperm were randomly assigned to create three categories: mated, pheromone-exposed or naïve. (1) For the mated category a male was placed in a drop with an adult female. Only males for which mating was confirmed by subsequent production of progeny were included in the results. (2) For the pheromoneexposed category a male was placed in a drop that had previously been occupied by an adult female. (3) For the naïve category a male was removed from and placed back into its own drop. After 48 h sperm development was recorded. Only males that had developed sperm in the mated and pheromone treatments (indicating successful exposure to female pheromone), and males without sperm in the naïve treatment (the condition of the majority of naïve males), were used in experiments. Males were photographed, and their area and length were measured using Image J 1.53e (http:// imagej.nih.gov/ij) as an index of their size.

Fighting Assays
Each male, including the controls, was transferred to a fresh drop of G. mellonella haemolymph using a sterile platinum wire and was allowed to acclimatize for 1 h. Males were then paired in drops symmetrically within category (mated with mated; pheromone-exposed with pheromone-exposed; naïve with naïve) or asymmetrically (mated with naïve; pheromoneexposed with naïve). There were thus five treatments. Control males of each category were taken out of and immediately replaced into their own drops. Pairs were observed for 30 min at 26 C. Fighting was recorded when one male wrapped around the other and tightened (Kapranas et al., 2020). Not all fights resulted in injury or death, and some pairs engaged in more than one fight during observation. Incidence of fighting, the identity of the initiator (in asymmetric pairs), number of fights, duration of fights and time at which fights occurred were recorded for each pair. A total of 33e59 pairs per treatment and 37e66 control individuals per category were observed. After 24 h at 20 C, mortality (including severe injury or paralysis) was recorded for each pair and control.

Data Analysis
The area and length of males from each category were compared using one-way analysis of variance (ANOVA) with significance at P < 0.05, followed by Tukey post hoc tests. For pairs in which fighting was observed, the number of fights, median duration and latency were analysed using ANOVA or the KruskaleWallis test, with significance at P < 0.05. For asymmetric fights, a one-sample proportion test was used to compare initiation of fights between conditioned (mated or pheromone-exposed) and naïve males with a hypothesized proportion initiated by naïve males of 0.5. The identity of the victim in these fights was also similarly tested for deviation from 0.5. To assess whether a mated male assesses its opponent, the probability of a mated male initiating an attack against a naïve male or against another mated male was compared using chi-square analysis. All the above data analysis was performed using Minitab 20.3 statistical software (Minitab, LLC, 2021) (or RStudio version 2022.12.0 (RStudio Team, 2020)). The incidence of fighting and death as influenced by different worm pairs (treatments) was explored with a binary logistic analysis. We used backward stepwise procedures and aggregation of factor levels (akin to a post hoc test) to obtain the parsimonious 'minimal adequate model' by model simplification . In the subsets of observations for which worm measurements were available for asymmetric pairings (pheromone-exposed versus naïve: N ¼ 24; mated versus naïve: N ¼ 19), we tested whether the initiation of fighting and the winning of fights by naïve worms was influenced by their opponent's status and the size (area) difference between them using a binary logistic regression . These analyses were performed with GENSTAT v. 22 (VSN International Ltd, Hemel Hempstead, U.K.). The same analysis could not be carried out for symmetrical pairings as the individuals were indistinguishable after pairing.

Ethical Note
No ethical approval was required for the species used in this research. The methodology used in this study was in keeping with the ASAB/ABS guidelines for the treatment of animals in behavioural research, as well as the suggestions by Drinkwater et al. (2019) for ethical treatment of invertebrates, with a focus on reduction and refinement. The number of G. mellonella used for haemolymph extraction was minimized by calculation of exact required volumes and the number of nematodes used in staged fights was reduced through refinement of the protocol.

Fighting and Mortality
There was a significant difference in incidence of fighting across treatments (logistic analysis: G 4 ¼ 3.04, P ¼ 0.016). Incidence of fighting was highest in symmetric mated pairs and lowest in symmetric pheromone-exposed pairs compared to all other treatments (Fig. 2) Table A1).
After 24 h, the overall incidence of damage or death in the pairs was 26% (61/234). There was no difference across treatments in incidence of damage or death in the pairs (logistic analysis: Table A1). Control mortality was low (3%; 10/300) and did not differ between categories of male (Pearson chi square: X 2 2 ¼ 1.8, P ¼ 0.405).

Identity of Initiator and Victim of Fighting in Asymmetric Pairings
When a fight occurred in asymmetric pairings, it was more likely to be the naïve male than the conditioned male that initiated fighting. This was consistent for both the mated versus naïve and pheromone-exposed versus naïve treatments (one-sample proportion test: X 2 1 ¼ 6.368, P ¼ 0.012 and X 2 1 ¼ 5.4, P ¼ 0.020, respectively; hypothesized proportion ¼ 0.5; Fig. 3, Appendix Table A2). In mated versus naïve pairs, where one male was damaged or dead after 24 h it was more likely to be the naïve male than the mated male (one-sample proportion test: X 2 1 ¼ 7.118, P ¼ 0.008; hypothesized proportion ¼ 0.5), but in the case of Area (mean ± SE) of S. carpocapsae males that were conditioned for 48 h in haemolymph drops with a female (mated), in drops previously occupied by a female (pheromone) and in their own drops (naïve). The number of males measured in each category is indicated on the bars. Bars with differing letters are significantly different (P < 0.001, Tukey's post hoc test). Measured males in the mated and pheromoneexposed categories all had sperm present while none of the naïve males had sperm. pheromone-exposed versus naïve males, there was no deviation from equality (one-sample proportion test: X 2 1 ¼ 0.474, P ¼ 0.491; hypothesized proportion ¼ 0.5; Fig. 4).
When mated versus naïve and pheromone-exposed versus naïve male pairs were compared, naïve males were as likely to initiate a fight when paired with a mated male as with a pheromone-exposed male (logistic analysis: G 1 ¼ 1.21, P ¼ 0.271). Analysis of the subsets for which worm size measurements were available confirmed that the probability of naïve worms initiating a fight was not influenced by competitor status (mated or pheromone-exposed) or size (opponent's status: G 1 ¼ 0.34, P ¼ 0.562; size: G 1 ¼ 1.45, P ¼ 0.228; interaction between opponent's status and size: G 1 ¼ 1.36, P ¼ 0.244).
Naïve males had a marginally higher but not significant tendency to win when paired against pheromone-exposed than when paired with mated males (logistic analysis: G 1 ¼ 3.31, P ¼ 0.069). However, in the subsets for which worm size measurements were available, competitor status (mated or pheromone-exposed) and size were not significant (opponent's status: G 1 ¼ 1.17, P ¼ 0.279; size: G 1 ¼ 1.22, P ¼ 0.269; interaction between opponent's status and size: G 1 ¼ 0.02, P ¼ 0.902).

DISCUSSION
As seen previously in S. longicaudum males (Kapranas et al., 2020), mated S. carpocapsae males had a killing advantage when paired against naïve males. Contrary to our hypothesis, however, this advantage could not be explained by the physiological effects induced by exposure to female pheromone, such as sperm production and the associated increase in size, since males that received pheromone-exposure treatment alone did not have a similar advantage over naïve males. Both mated and pheromoneexposed males were similar in size, being larger than naïve males, presumably due to the sperm stored in the seminal vesicle    increasing the worms' diameter. Despite the similarity in size, mated and pheromone-exposed males differed in the outcome of interactions with naïve males, in that mated but not pheromonetreated males were more likely to be the victor over the naïve male. Moreover, analyses on asymmetric pairs showed that the size difference between opponents did not affect the outcome of the interaction. Body size is an important determinant of RHP: in general, larger animals are more likely to win (Archer, 1988;Brown et al., 2006;Hsu et al., 2006;Hughes, 1996) but this is not always the case. For example, Elwood et al. (1998) found that relative size affected neither the cost nor the probability of victory in fighting in the hermit crab Pagurus bernhardus. Although both the mated and pheromone-exposed males had developed sperm and increased in diameter, they differed in their experience with a female. In many animal taxa, previous mating has been shown to increase a male's self-assessment of RHP, thus increasing their confidence and likelihood of fighting (Guevara-Fiore et al., 2012; Killian & Allen, 2008;Kralj-Fi ser et al., 2011;Yasuda et al., 2015). This is similar to the 'winner effect': animals that have succeeded in combat are more likely to succeed again (Hsu et al., 2006;Kasumovic et al., 2009;Kou & Hsu, 2013). Experience gained during mating may also increase skills that contribute to an advantage in fighting (Briffa & Lane, 2017). In Steinernema, the male wraps its tail end around the female at the vulva and uses a pair of copulatory spicules to assist insemination (Lewis et al., 2002). This behaviour resembles the coiling used in fighting, and the spicules may assist in inflicting injuries either by puncturing or by focusing the pressure delivered by squeezing (Zenner et al., 2014). Mating behaviour of male nematodes is controlled by a specialized posteriorly located 'connectome', a network of neurons, muscles and gonad (Jarrell et al., 2012) that may also have a role in fighting behaviour in Steinernema. In Caenorhabditis elegans, mating experience results in the rewiring of specific synapses in the male connectome, with evidence that these changes are translated into finer coordination of muscle contraction, such as those involved in spicule protraction (Hart & Hobert, 2018).
In our experiments, we exposed males to female pheromone to induce physiological maturation, to explore the importance of the mated male's physical status in winning a fight. In other animals, experience of a female (without mating) may or may not mimic the effects of actual copulation on male aggression (Bergman et al., 2010;Judge et al., 2010;Killian & Allen, 2008;Rillich et al., 2019). For example, male crickets, A. domesticus, that were allowed to contact a female but not to copulate with her showed the same increased aggression as males that were allowed to copulate, suggesting that chemotactile cues from the female were sufficient to elicit this change in aggression (Killian & Allen, 2008). While female pheromone gives an indication to S. carpocapsae males of female presence, it is unclear whether it can be considered equivalent to the 'female experience' treatment of studies in other animals. In nematodes, the response of males to female pheromone represents the first step in mate finding, but there is also a second step involving contact cues at the female's body surface (Barr & Garcia, 2006;Sakai et al., 2013).
Pheromones are widely implicated in modulating aggression in insects and mammals (Itakura et al., 2022;Sengupta et al., 2022). In nematodes, pheromones are involved in numerous social behaviours including mate attraction, aggregation and dispersal (Edison, 2009;Hartley et al., 2019;Muirhead & Srinivasan, 2020). Nematode pheromones mainly belong to a class of molecules called ascarosides, a modular library of potent molecules with both developmental and behavioural effects (Choe et al., 2012;Butcher, 2017;Park et al., 2019). Media conditioned by S. carpocapsae females contain a blend of ascarosides and both attract conspecific males and initiate their gonadal maturation (Choe et al., 2012;Hartley, 2017). Exposure to female pheromone in our experiments clearly induced the physical effects (sperm development and concomitant increase in size) but not the behavioural effects of mating. Indeed, pairs of pheromone-treated males showed the lowest incidence of fighting, in contrast to the elevated fighting in pairs of mated males. Female steinernematids have a short window of availability, becoming less attractive (Lewis et al., 2002;Hartley, 2017) and of lower RV (Kapranas et al., 2020) for males shortly after mating. In the closed environment of a dead insect, residual pheromone, in the absence of available females, would indicate that there were no more available females in the founding generation, and the only way for a male to achieve at least some reproductive success would be to survive until females of the filial generation become available (Zenner et al., 2014). Thus, the reduced aggression of pheromone-matured males would be adaptive. Internal state, social interaction and other experiences interact in a complex manner to influence the translation of stimuli into behaviours, even in a relatively simple organism such as a nematode (Hashikawa et al., 2018;Park et al., 2019). The impact on a male steinernematid of experiencing pheromone only, without encountering a female, warrants further exploration.
Although mated males tended to win fights against naïve males, it was the naïve male that was more likely to initiate fighting in these pairings. Naïve males are expected to value a mating opportunity more highly than a male that has just mated (Kemp, 2006) and will be more aggressive as a result (Brown et al., 2006(Brown et al., , 2007. This may be a case of contradictory asymmetry where the opponent with the lower RHP has more to gain from the fight (Parker & Rubinstein, 1981). It is expected that if males can assess relative RHP, the male with lower RHP will terminate the fight before sustaining injury. This is based on mutual assessment (Enquist & Leimar, 1987). However, if the male with the lower RHP is in a position where it cannot reproductively succeed by quitting the fight then the 'desperado effect' comes into play and the male will continue to attack due to the lower potential cost (Grafen, 1987). Naïve males in this case exist in a divisive asymmetry where they will always lose out to mated males and pheromone-exposed males for mating opportunities. This is because naïve males have not developed sperm and are not ready to mate. By the time the naïve males are sexually developed the mature males may have taken all the available females. Therefore, it is to the naïve male's advantage to always risk the fight. Pheromone-exposed males do not have the same disadvantage as they are sexually developed and are therefore less motivated to fight and risk their future reproductive success.
Comparison of symmetric and asymmetric pairings involving mated males suggests that males are able to assess their opponents' state (using physiological and/or behavioural clues) and thus their relative RHP; mated males were three times more likely to initiate fighting in matedemated than in matedenaïve pairings. This may be related to the threat that the opponent poses as a competitor in reproduction. A mated male will risk attacking another mated male, as it is ready to mate, but may decide that attacking a naïve male is not worth it, as it is not ready to mate. Despite the high incidence of fighting in matedemated pairs relative to all other pairings, the mortality rate was not similarly elevated in this treatment. Fighting should end when one male has died or gives up the attack. When opponents are equally matched the cost of fighting is highest due to continued escalation (Arnott & Elwood, 2009;Payne and Pagel, 1996). If males can assess each other's RHP they may decide to terminate fights before injury occurs when equally matched (Enquist & Leimar, 1990).
Adding to the previously described effects of competitor density and relatedness (Zenner et al., 2014;Kapranas et al., 2016), residency and objective RV (Kapranas et al., 2020), this study further shows the value of the entomopathogenic nematodes in testing theoretical predictions in contest behaviour and in behavioural ecology in general (Lewis et al., 2022).
In conclusion, there are considerable asymmetries due to mating status between Steinernema males in conflict behaviour, in their tendency both to initiate and to win a fight. The advantage of mated Steinernema males over naïve opponents cannot be explained by larger size, in contrast to findings reported for many animal taxa, but may be due to other elements of RHP or RV.

Data Availability
Data are available as Supplementary material.

Declaration of Interest
None.