Males in many species possess sexually selected weapons that they use to fight for mating opportunities. It is well established that male-male competition can lead to physical injuries for males. However, very few studies have looked at the physical consequences for conspecific females. We hypothesized that living with males in a species with male-male competition would result in female injury. Because larger female invertebrates typically have greater reproductive output, they have higher resource value for males and can elicit aggression and fighting. Thus, we further hypothesized that larger females in this context would receive more injuries. For this study, we focused on the leaf-footed cactus bug, Narnia femorata (Hemiptera: Coreidae), a species of insect in which males fight using their spiny and enlarged hindlegs. In just two hours of observation, we documented males competing with other males in 61% of 103 trials. In 43% of these 63 competitions, females were physically contacted and sometimes attacked with a kick or squeeze. We left insects in social groups for 74 hours and found that females living with multiple males had a higher likelihood of obtaining injuries (26.2% of 103 trials) compared to those living only with females (9.7% of 103 trials). In addition, larger females were more likely to be injured compared to smaller females. Our study highlights the harm that females can experience in species with male-male competition.

Insect rearing

We used Narnia femorata (Stål, 1892) (Hemiptera: Coreidae) from a lab colony established in 2022 and supplemented in 2023 with wild-caught individuals from Live Oak, Florida (30.2642 N°, 83.1768° W). All the insects were kept in a rearing room with temperatures ranging from 26–28 °C. We reared nymphs in plastic deli cups capped with mesh lids. We kept the density at 5-20 siblings per cup since nymphs are usually found in groups in nature (Allen and Miller 2020). We also provided each cup with a prickly pear cactus pad (Opuntia mesacantha ssp. lata) planted in soil with an adjacent cactus fruit. We monitored the quality of cacti and cactus fruit throughout the project and moved bugs to a new cup if cactus quality declined.

When the nymphs emerged into their penultimate instar, we separated each one into an individual deli cup with a planted cactus pad and a fruit. They were kept in deli cups by themselves to minimize social interactions and to ensure the females remained unmated. Once an individual emerged as an adult, we assigned it a unique code and marked it using a non-toxic paint pen. All adults were marked at least one day before the experiments started to minimize any effects of this disruption during the experimental period.

Treatments

There were two treatments, each with 103 replicates: a 1) female-only treatment and a 2) mixed-sex treatment. Each group of the female-only treatment consisted of one focal female and three background females; each group of the mixed-sex treatment consisted of one focal female and three males. For the mixed-sex treatment, we mimicked an operational sex ratio where more males are seeking matings than are females, a situation likely common in nature (Kokko et al. 2014). The distribution of these insects in the wild suggests the formation of hidden leks (sensu Fletcher & Miller 2006), and it is also not uncommon in the leaf-footed bug superfamily to find multiple males on a host plant as they vie for a coveted territory (Miyatake 1995). Not considering reproductive state or social groupings, the primary sex ratio of N. femorata in the wild is typically 1:1 (Cirino and Miller 2017). Females mate multiply, and work on other hemipteran insects suggests that last-male sperm precedence is common (Parker 1970; Gwynne 1984).

We assigned insects randomly to a treatment group. Some unmated females were used as background females in two different replicates of the female-only treatment. Each male was used in the mixed-sex treatment up to five times. We visually inspected each focal insect before assigning it to a treatment and did not include those with obvious physical damage.

Experimental Design

There were two parts of the experiment: a two-hour observation period followed by a 72-hour non-observation period. During the observation period, we conducted behavioral trials at approximately 29 °C, noting their fighting and copulating behaviors. Before each trial, we allowed selected insects to acclimate to the environment in mesh-covered plastic cups alone for around 30-60 minutes before putting them into their assigned group. All insects were gently moved into their assigned deli cup with a cactus and a ripened fruit. We also placed pine needles into the container, which females used as their main egg-laying substrate.

Each observation period lasted for 2 hours, and we watched up to 9 groups at a time. It was feasible to track behaviors for multiple groups at the same time because these insects interact only intermittently and sometimes appear to ignore each other completely (Nolen et al. 2017). We defined male-male competition (or male contest) as one or more behaviors that include the following: leg displaying, charging, mounting, kicking, or squeezing. We defined these behaviors as in the ethogram (Table S1) modified from Nolen et al. (2017). We conducted observation on these behaviors for both mixed-sex and female-only treatments. The squeeze attack, where males squeeze each other using their spiny hindlegs, and often end-to-end, is part of male-male competition, and is the behavior most likely to cause injury because of the sharp spines and the force generated as the tibia and femur of one or both legs are squeezed together. In fact, squeezing can be felt as a pinch and can even draw blood when some leaf-footed bugs are handled by humans (Miller et al. 2024). Thus, we paid special attention to hindleg squeezes in one of our statistical analyses. Copulation was noted when a male mounted a female, initiated genital contact, and females opened their genital plates to allow copulation. Copulation was only considered successful when a male subsequently dismounted and rotated outward to form an obtuse angle with the genitalia still connected. We documented the number of copulating pairs that were contacted by rivals and whether a kick or squeeze occurred.

After the two-hour observation period, we moved all the groups back to the 26-28 °C rearing room for another 72 hours. We provided this extended period to allow insects to interact further and females in the mixed-sex treatment to lay eggs so that we could quantify their reproductive output. Females in the female-only treatment were always unmated. After this non-observation period, we froze the focal female in each group for both treatments for later measurement and bodily injury scoring.

Injury Scoring

We scored female injury blind to treatment using a Leica M165C stereo microscope. We used ImageJ v.1.54d (Schneider et al. 2012) to measure pronotum width as a proxy for body size (Allen and Miller 2017). We counted and categorized observed physical injuries as follows: (1) antennal injury, including broken antennae and missing setae; (2) leg injury, including broken spines, melanization, exoskeletal punctures, and missing legs; and (3) wing injury, including wing tears and wing fragmentation. We noted evidence of melanization because it suggests wound healing in insects (Gillespie et al. 1997).

Offspring production

We assessed one potential negative consequence of injury – a decrease in the production of offspring. We only collected these reproductive output measurements for the mixed-sex treatment groups because females remained unmated in the female-only groups. We assessed the number of live offspring produced from eggs laid during the 74-hour experimental period. We counted the number of hatchlings based on the number of opened pseudo-opercula on Day 14 after the experiment. Hatchling counts were collected by two people, blind to each other’s counts, and we used the average of these two numbers.

Statistical Analyses

All statistical analyses were conducted using IBM SPSS v.28. We first tested the consequence of the mixed-sex versus the female-only treatment on female injury after the 74-hour experimental period. We used a Generalized Linear Model (GLM) assuming a binominal distribution with female injury (presence or absence) as a binary response variable. The explanatory variable was social group treatment (mixed-sex or female-only), which was considered as the main effect. We used female body size as a continuous covariate. In this and in all models, we first included two-way interaction. We a priori decided to use the stepwise elimination procedure (Hardy and Field 1998), where non-significant interactions are removed. We conservatively decided a priori that only interactions with p > 0.10 would be removed.

Next, we examined whether an observed aggressive squeeze attack on a female was predictive of female injury at the end of the 74 hours. We constructed a binomial GLM with female injury (presence or absence) as the binary response variable. Our model tested if an observed squeeze attack on a female predicted female injury at the end of the study. This GLM included female body size as a continuous covariate.

Finally, we examined the effects of female body size and injury on offspring production. The hatchling count data was over-dispersed, thus we ran a GLM fit to the negative binomial distribution. Our model included the number of hatchlings as the response variable and both female injury (presence or absence) and female body size as explanatory variables.