Periodic Reporting for period 1 - SCISM (Sexual conflict in the sexually cannibalistic Springbok mantis (Miomantis caffra))
Reporting period: 2023-09-01 to 2025-08-31
Summary of the context and overall objectives of the project
Conflict between the sexes is widespread in nature and occurs when the evolutionary interests of males and females are different. Sexual cannibalism—which occurs when a female eats her mate before, during or after mating—is considered a classic example of sexual conflict since females gain a meal but males lose their life. The risk of premature death is expected to select for cautious mating tactics in males that help them to secure matings and avoid being eaten. Although such stealthy mating tactics are common in cannibalistic species, males of the Springbok mantis, Miomantis caffra, take a far more aggressive approach, violently wrestling females to coerce them to mate. Males that win wrestling bouts are more likely to secure matings and avoid cannibalism, but wrestling severely wounds some females whose abdomens are punctured by the pointed claws of males’ raptorial forelegs that are usually used for catching prey. While this harmful male behaviour appears to be a classic manifestation of sexual conflict, the exact causes and consequences of male coercion and female injury in this species are yet to be elucidated. My project aims to use field, laboratory, and modelling approaches to investigate the role of sexual conflict in the evolution of cannibalistic and anti-cannibalistic behaviour. First, to determine rates of female injury in the wild, I plan to use mark-recapture techniques to track sex-ratio and injury incidence through time in several wild populations of M. caffra in Auckland, New Zealand, where the species is a naturalised pest. Second, I plan to run a series of laboratory experiments to explore whether, as sexual conflict theory predicts, males increase their fitness by injuring females, or whether injury is an incidental side-effect of males’ drive to mate. Third, I plan to run a mesocosm experiment in the university’s glasshouse facilities to examine whether an increase in the intensity of sexual conflict reduces female survival and population productivity—a key theoretical prediction that has rarely been tested. Finally, I intend to use modelling techniques to investigate, more generally, the role that sexual conflict over mating decisions plays in shaping male and female behaviour in sexually cannibalistic taxa—a relationship that is currently poorly understood.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
The following is an assessment of the completion of the different components of the original proposal’s work plan (see Table 2, p. 9). Each laboratory experiment (WP2, WP3, and WP4) mesocosm experiment (WP6), and field study (WP5) was carried out successfully as described in the proposal. However, following consultation with my collaborator and secondment host, the methodology for the theoretical study (WP1) was altered slightly to utilise a game-theoretical approach in conjunction with the originally proposed individual-based models. All projected milestones relating to data collection and analysis were reached for M1, M2, M4, M5 and M6. Data has been collected for M3 but analysis is still underway. For each of the above work packages (except WP1), a preregistered research plan was submitted to osf.io. These deliverables can be openly accessed at the following links: D2.1: https://osf.io/z5ce2(opens in new window) D3.1: https://osf.io/7g26b(opens in new window) D4.1: https://osf.io/srtvn(opens in new window) D5.1: https://osf.io/svha4(opens in new window) D6.1: https://osf.io/hg3tz(opens in new window). The preregistration research plan for WP1 (i.e. D1.1) will be uploaded to osf.io in the coming weeks. While all data has been collected and analysed (except for the analysis for WP3), to date only one of the 6 work packages (WP2) has been written up into a submitted manuscript. The draft manuscript (D2.2) is freely available as a preprint on BioRxiv.org (https://www.biorxiv.org/content/10.1101/2025.08.21.671560v1(opens in new window)) and has been submitted for open-access publication at Royal Society Open Science, where it is currently under review. The data underpinning this study will be made publicly available through the data repository datadryad.org upon publication. The deliverables for the other work packages (i.e. D1.2 D1.3 D2.3 D3.2 D3.3 D4.2 D4.3 D5.2 D5.3 D6.2 D6.3) are all currently in progress and will be completed in the coming months.
A data management plan (WP7) has been submitted with regular updates, and results have been presented at local and international conferences (WP8), namely: Congress of the International Society for Behavioural Ecology, Melbourne, Australia, 2024; and Ethological Society Conference, Münster, Germany, 2024. Dissemination beyond publications and conferences has taken different forms that were not originally planned for. I was unable to enter any of the photography competitions I originally planned for (i.e. D9A, D9B, D9C, D9D) because I ran out of research funds to by appropriate photography equipment. But I gave a pop-science talk about my research as part of the Pint of Science Festival in Hamburg, Germany in 2024, and participated in a pop-science interview for the podcast I Came For Science in 2025 (https://open.spotify.com/episode/23KIlUtosGX74tQly0GgAt?si=6051a40ae0394984(opens in new window)). There are additional dissemination deliverables in the works as well. I intend to send out press releases to the media (i.e. D9.1 D9.2 D9.3 D9.4 D9.5 D9.6) upon publication of the rest of the above-listed manuscripts. I am also still in the process of collaborating with my artist friends to produce a short film about my research (i.e. D9.7) which will be released in later months.
I was able to participate in extensive training (WP10) over the course of my fellowship. I demonstrated and tutored (M10.2) in two Master’s courses (MBIO-AB-2 Evolution and Behaviour; MBIO-AB-2 Behavioural Ecology), and designed and delivered my own Master’s course in data management (MBIO-W-57 Introduction to R for Master’s Students) for two semesters at the host institute (UHH). I participated in two German-language courses (10.3) attaining level A2. I also enrolled in a grant-writing skills workshop (M10.4) but was unable to attend due to illness. Other more unstructured training (M10.1) included supervising 5 bachelor student projects and one Master’s, providing advice to other Master’s and PhD students on their experimental design and statistics, and directing my research budget with guidance from my academic host. A career development plan (D10) has been submitted with regular updates.
A data management plan (WP7) has been submitted with regular updates, and results have been presented at local and international conferences (WP8), namely: Congress of the International Society for Behavioural Ecology, Melbourne, Australia, 2024; and Ethological Society Conference, Münster, Germany, 2024. Dissemination beyond publications and conferences has taken different forms that were not originally planned for. I was unable to enter any of the photography competitions I originally planned for (i.e. D9A, D9B, D9C, D9D) because I ran out of research funds to by appropriate photography equipment. But I gave a pop-science talk about my research as part of the Pint of Science Festival in Hamburg, Germany in 2024, and participated in a pop-science interview for the podcast I Came For Science in 2025 (https://open.spotify.com/episode/23KIlUtosGX74tQly0GgAt?si=6051a40ae0394984(opens in new window)). There are additional dissemination deliverables in the works as well. I intend to send out press releases to the media (i.e. D9.1 D9.2 D9.3 D9.4 D9.5 D9.6) upon publication of the rest of the above-listed manuscripts. I am also still in the process of collaborating with my artist friends to produce a short film about my research (i.e. D9.7) which will be released in later months.
I was able to participate in extensive training (WP10) over the course of my fellowship. I demonstrated and tutored (M10.2) in two Master’s courses (MBIO-AB-2 Evolution and Behaviour; MBIO-AB-2 Behavioural Ecology), and designed and delivered my own Master’s course in data management (MBIO-W-57 Introduction to R for Master’s Students) for two semesters at the host institute (UHH). I participated in two German-language courses (10.3) attaining level A2. I also enrolled in a grant-writing skills workshop (M10.4) but was unable to attend due to illness. Other more unstructured training (M10.1) included supervising 5 bachelor student projects and one Master’s, providing advice to other Master’s and PhD students on their experimental design and statistics, and directing my research budget with guidance from my academic host. A career development plan (D10) has been submitted with regular updates.
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
The funded project generated fascinating results that contribute substantially to our understanding of the evolution of sexual cannibalism by sexual conflict. Because these findings have not yet been peer-reviewed or formally published, they should be considered preliminary.
Work package WP1 used game-theoretical and individual-based models to investigate the conditions under which sexual conflict can promote the invasion and evolutionary maintenance of sexual cannibalism as a female resistance trait. The models showed that when there is sexual conflict over mating number, non-copulatory cannibalism (that is, cannibalism occurring without any mating taking place) is always selectively favoured over other non-cannibalistic forms of resistance because it provides a dual benefit to females of avoiding mating costs and boosting nutrition. However, cannibalism can only evolve in the absence of conflict if the nutritional benefit of eating males is extremely high. The models additionally identified the best cannibalism-vs-mating strategy for optimising fitness (which was: “mate first, cannibalise thereafter”), and identified conditions under which more maladaptive strategies that utilise cannibalism in the first sexual encounter can be maintained by stabilising selection, resulting in stable polymorphisms between invasive cannibalistic strategies and wildtype non-cannibalistic strategies. An extension of the model showed that sexual cannibalism can be strongly favoured by protandry, which occurs when males reach maturity faster than females. When development is protandrous, a greater range of noncopulatory cannibalistic strategies can be maintained indefinitely in temporal polymorphism with non-cannibalistic wildtype strategies, such that cannibalistic females proliferate early in the season when males are abundant and non-cannibalistic females are favoured later in the season when males are rarer. Additional supplementary models were developed to investigate the evolution of sexual cannibalism during mating in response to sexual conflict over the duration of mating, and broadly similar findings were obtained. Overall, these results highlight the importance of sexual conflict in the evolution of sexually cannibalistic behaviour, providing the first theoretical basis for the verbal hypothesis that cannibalism functions as a female resistance trait. The models also provide exciting new predictions, such as the idea that sexual conflict can generate stable polymorphisms in wild populations of sexually cannibalistic taxa.
Work package WP2 investigated whether extra-genital injury inflicted on females by males has evolved because it is (1) adaptive per se for males, or (2) a pleiotropic side-effect of selection on other male traits that are adaptive in other contexts. Unlike other systems where sexual injury occurs internally via weaponised genitalia, males of the springbok mantis (Miomantis caffra) stab females in the abdomen with their foretibial claws while fighting back against sexually cannibalistic attacks. Females are often seriously wounded by these encounters, resulting in haemolymph loss and even prolapse of the intestines through the stab wounds. If injury is adaptive, then experimentally injured females should alter their behaviour to benefit their mates by becoming less attractive or more cannibalistic toward second males, or by delaying remating or remating less. In addition, males could benefit by inducing females into a terminal investment response, causing them to die faster or produce more eggs or offspring sooner. The results for this phenotypic engineering experiment found little evidence for adaptive harm. Injured females did not alter their attractiveness, levels of cannibalism, or remating likelihood compared to intact females. Injured females also showed no change in mortality, fecundity or offspring production. However, there was a significant delay in the remating period among injured females, which could potentially represent a benefit to males if such a delay reduces sperm competition. Whether longer remating periods also occur in more naturalistic conditions remains to be tested. These results suggest that injury in M. caffra is likely to be an incidental side-effect of males attempting to avoid sexual cannibalism. This work additionally suggests that extra-genital injury could be a fruitful direction for future research on the evolution of sexual harm.
Work package WP3 investigated the factors that correlate with mating, sexual cannibalism and female injury in the springbok mantis (Miomantis caffra). Data from this observational study have been collected but are yet to be analysed, so results are not available.
Work package WP4 investigated the function of foretibial claws in the springbok mantis (Miomantis caffra). Foretibial claws occur in both male and female mantises and have evolved primarily for prey capture. However, since M. caffra males also use their claws for grasping hold of females during premating struggles that they must win in order to mate, male claws may have been co-opted for optimising mating outcomes as well. If claws are important for prey capture, then males with intact claws should be better than males with no claws at catching and/or handling prey. If claws also function to facilitate mating, then males with intact claws compared to males with no claws should be more likely to mate and be faster at securing a mating. Results from this phenotypic engineering experiment demonstrate that claws do not significantly alter males’ ability to catch or handle prey, since males with and without claws did not differ in the time taken to capture prey or the time spent consuming prey. However, the presence/absence of claws interacted with other traits to shape mating success. Females originating from laboratory stocks were more likely to mate with clawless males than intact males, and larger males were more likely to mate when clawless than when intact, suggesting that males’ motivation to mate is enhanced in certain contexts by the experimental removal of their claws. This interpretation is further supported by the fact that clawless males mated sooner when paired with larger females. Curiously, these results suggest that foretibial claws may play a more important role in male mating decisions than in male hunting success, but not in the direction predicted, with clawless males potentially mating faster as a terminal response to injury. Future work could additionally assess whether claws facilitate easier capture of prey of different sizes.
Work package WP5 investigated the presence and seasonality of female injury and cannibalism in a longitudinal field study in wild populations of the springbok mantis (Miomantis caffra). As mentioned above, females of this species are highly aggressive and cannibalistic towards males, but males fight back and injure females by stabbing them in the abdomen during premating struggles that leave behind distinctive scabs over the stab wounds. Injury is expected to be more likely in heavier/fatter females, since the abdomens of such females should be plumper and easier for males to stab. Injury is therefore expected to increase in frequency as the season progresses and as female body weight increases through time. Moreover, because high rates of noncopulatory cannibalism have been observed in the laboratory, females are also expected to cannibalise in the wild, which should be observable in the rapid weight gain of individual females or the reduced abundance of males. The results from this mark-recapture study indicate that injury is uncommon in the wild: no cases of injury were observed in any female over the three-month period of the study. However, the sex ratio was persistently female-biased among adults, but male-biased among juveniles, indicative of a decline in male numbers due to sexual cannibalism. There was additional suggestive evidence of cannibalism when comparing the weight gain of wild females to that of laboratory females. Taken together, these results suggest that M. caffra females have the upper hand over males in sexual interactions in the wild, with sexual cannibalism appearing to occur commonly. Future research could assess whether the absence of injury in the wild is due to males being more choosy about when and who to mate with.
Work package WP6 investigated the effect of male abundance on the manifestation of sexual conflict and population productivity in the springbok mantis (Miomantis caffra). Sexual conflict theory predicts that males and females can have divergent interests for certain mating outcomes, with conflict becoming more intense and females suffering higher costs (such as higher mortality or more severe sexual injury) when males are more abundant. However, in taxa where females cannibalise males without mating and gain nutritional benefits by doing so, an increase in conflict due to a higher abundance of males could lead to a greater incidence of cannibalism and therefore enhanced female fitness. The results from this mesocosm experiment that manipulated male abundance across multiple controlled populations indicate that higher male abundance increases the incidence, but not the rate, of sexual cannibalism. Females exposed to more abundant males incurred injuries that were more severe, probably because of their higher likelihood of mating. But this greater severity did not result in higher costs, since females exposed to more abundant males did not die faster and produced a similar number of offspring compared to females exposed to less abundant males. These results indicate that females respond to excess males with cannibalism and suffer more severe injuries by interacting with them, but the greater resources obtained by eating these males does not translate to a net fitness gain for such females or a net fitness loss due to injury. This suggests that, as predicted by sexual conflict theory, females use sexual cannibalism to control mating outcomes, but that male-imposed costs drive a compromise in fitness outcomes. These results also suggest that sexual cannibalism is unlikely to be an optimal foraging strategy, since females attacked and consumed males at high rates despite having access to an abundance of other prey.
Work package WP1 used game-theoretical and individual-based models to investigate the conditions under which sexual conflict can promote the invasion and evolutionary maintenance of sexual cannibalism as a female resistance trait. The models showed that when there is sexual conflict over mating number, non-copulatory cannibalism (that is, cannibalism occurring without any mating taking place) is always selectively favoured over other non-cannibalistic forms of resistance because it provides a dual benefit to females of avoiding mating costs and boosting nutrition. However, cannibalism can only evolve in the absence of conflict if the nutritional benefit of eating males is extremely high. The models additionally identified the best cannibalism-vs-mating strategy for optimising fitness (which was: “mate first, cannibalise thereafter”), and identified conditions under which more maladaptive strategies that utilise cannibalism in the first sexual encounter can be maintained by stabilising selection, resulting in stable polymorphisms between invasive cannibalistic strategies and wildtype non-cannibalistic strategies. An extension of the model showed that sexual cannibalism can be strongly favoured by protandry, which occurs when males reach maturity faster than females. When development is protandrous, a greater range of noncopulatory cannibalistic strategies can be maintained indefinitely in temporal polymorphism with non-cannibalistic wildtype strategies, such that cannibalistic females proliferate early in the season when males are abundant and non-cannibalistic females are favoured later in the season when males are rarer. Additional supplementary models were developed to investigate the evolution of sexual cannibalism during mating in response to sexual conflict over the duration of mating, and broadly similar findings were obtained. Overall, these results highlight the importance of sexual conflict in the evolution of sexually cannibalistic behaviour, providing the first theoretical basis for the verbal hypothesis that cannibalism functions as a female resistance trait. The models also provide exciting new predictions, such as the idea that sexual conflict can generate stable polymorphisms in wild populations of sexually cannibalistic taxa.
Work package WP2 investigated whether extra-genital injury inflicted on females by males has evolved because it is (1) adaptive per se for males, or (2) a pleiotropic side-effect of selection on other male traits that are adaptive in other contexts. Unlike other systems where sexual injury occurs internally via weaponised genitalia, males of the springbok mantis (Miomantis caffra) stab females in the abdomen with their foretibial claws while fighting back against sexually cannibalistic attacks. Females are often seriously wounded by these encounters, resulting in haemolymph loss and even prolapse of the intestines through the stab wounds. If injury is adaptive, then experimentally injured females should alter their behaviour to benefit their mates by becoming less attractive or more cannibalistic toward second males, or by delaying remating or remating less. In addition, males could benefit by inducing females into a terminal investment response, causing them to die faster or produce more eggs or offspring sooner. The results for this phenotypic engineering experiment found little evidence for adaptive harm. Injured females did not alter their attractiveness, levels of cannibalism, or remating likelihood compared to intact females. Injured females also showed no change in mortality, fecundity or offspring production. However, there was a significant delay in the remating period among injured females, which could potentially represent a benefit to males if such a delay reduces sperm competition. Whether longer remating periods also occur in more naturalistic conditions remains to be tested. These results suggest that injury in M. caffra is likely to be an incidental side-effect of males attempting to avoid sexual cannibalism. This work additionally suggests that extra-genital injury could be a fruitful direction for future research on the evolution of sexual harm.
Work package WP3 investigated the factors that correlate with mating, sexual cannibalism and female injury in the springbok mantis (Miomantis caffra). Data from this observational study have been collected but are yet to be analysed, so results are not available.
Work package WP4 investigated the function of foretibial claws in the springbok mantis (Miomantis caffra). Foretibial claws occur in both male and female mantises and have evolved primarily for prey capture. However, since M. caffra males also use their claws for grasping hold of females during premating struggles that they must win in order to mate, male claws may have been co-opted for optimising mating outcomes as well. If claws are important for prey capture, then males with intact claws should be better than males with no claws at catching and/or handling prey. If claws also function to facilitate mating, then males with intact claws compared to males with no claws should be more likely to mate and be faster at securing a mating. Results from this phenotypic engineering experiment demonstrate that claws do not significantly alter males’ ability to catch or handle prey, since males with and without claws did not differ in the time taken to capture prey or the time spent consuming prey. However, the presence/absence of claws interacted with other traits to shape mating success. Females originating from laboratory stocks were more likely to mate with clawless males than intact males, and larger males were more likely to mate when clawless than when intact, suggesting that males’ motivation to mate is enhanced in certain contexts by the experimental removal of their claws. This interpretation is further supported by the fact that clawless males mated sooner when paired with larger females. Curiously, these results suggest that foretibial claws may play a more important role in male mating decisions than in male hunting success, but not in the direction predicted, with clawless males potentially mating faster as a terminal response to injury. Future work could additionally assess whether claws facilitate easier capture of prey of different sizes.
Work package WP5 investigated the presence and seasonality of female injury and cannibalism in a longitudinal field study in wild populations of the springbok mantis (Miomantis caffra). As mentioned above, females of this species are highly aggressive and cannibalistic towards males, but males fight back and injure females by stabbing them in the abdomen during premating struggles that leave behind distinctive scabs over the stab wounds. Injury is expected to be more likely in heavier/fatter females, since the abdomens of such females should be plumper and easier for males to stab. Injury is therefore expected to increase in frequency as the season progresses and as female body weight increases through time. Moreover, because high rates of noncopulatory cannibalism have been observed in the laboratory, females are also expected to cannibalise in the wild, which should be observable in the rapid weight gain of individual females or the reduced abundance of males. The results from this mark-recapture study indicate that injury is uncommon in the wild: no cases of injury were observed in any female over the three-month period of the study. However, the sex ratio was persistently female-biased among adults, but male-biased among juveniles, indicative of a decline in male numbers due to sexual cannibalism. There was additional suggestive evidence of cannibalism when comparing the weight gain of wild females to that of laboratory females. Taken together, these results suggest that M. caffra females have the upper hand over males in sexual interactions in the wild, with sexual cannibalism appearing to occur commonly. Future research could assess whether the absence of injury in the wild is due to males being more choosy about when and who to mate with.
Work package WP6 investigated the effect of male abundance on the manifestation of sexual conflict and population productivity in the springbok mantis (Miomantis caffra). Sexual conflict theory predicts that males and females can have divergent interests for certain mating outcomes, with conflict becoming more intense and females suffering higher costs (such as higher mortality or more severe sexual injury) when males are more abundant. However, in taxa where females cannibalise males without mating and gain nutritional benefits by doing so, an increase in conflict due to a higher abundance of males could lead to a greater incidence of cannibalism and therefore enhanced female fitness. The results from this mesocosm experiment that manipulated male abundance across multiple controlled populations indicate that higher male abundance increases the incidence, but not the rate, of sexual cannibalism. Females exposed to more abundant males incurred injuries that were more severe, probably because of their higher likelihood of mating. But this greater severity did not result in higher costs, since females exposed to more abundant males did not die faster and produced a similar number of offspring compared to females exposed to less abundant males. These results indicate that females respond to excess males with cannibalism and suffer more severe injuries by interacting with them, but the greater resources obtained by eating these males does not translate to a net fitness gain for such females or a net fitness loss due to injury. This suggests that, as predicted by sexual conflict theory, females use sexual cannibalism to control mating outcomes, but that male-imposed costs drive a compromise in fitness outcomes. These results also suggest that sexual cannibalism is unlikely to be an optimal foraging strategy, since females attacked and consumed males at high rates despite having access to an abundance of other prey.