Skip to main content
NCBI home page
Proceedings of the Royal Society B: Biological Sciences logoLink to Proceedings of the Royal Society B: Biological Sciences
. 2021 Jan 27;288(1943):20202898. doi: 10.1098/rspb.2020.2898

Weapon performance drives weapon evolution

Zachary Emberts 1,✉, Wei Song Hwang 2, John J Wiens 1
PMCID: PMC7893261  PMID: 33499793

Abstract

Many sexually selected traits function as weapons, and these weapons can be incredibly diverse. However, the factors underlying weapon diversity among species remain poorly understood, and a fundamental hypothesis to explain this diversity remains untested. Although weapons can serve multiple functions, an undeniably important function is their role in fights. Thus, a crucial hypothesis is that weapon diversification is driven by the evolution of weapon modifications that provide an advantage in combat (e.g. causing more damage). Here, we test this fighting-advantage hypothesis using data from 17 species of coreid bugs. We utilize the fact that male–male combat in coreids often results in detectable damage, allowing us to link different weapon morphologies to different levels of damage among species. We find that certain weapon morphologies inflict much more damage than others, strongly supporting the fighting-advantage hypothesis. Moreover, very different weapon morphologies can inflict similarly severe amounts of damage, leading to a weapon performance landscape with multiple performance peaks. This multi-peak pattern could potentially drive different lineages towards divergent weapon forms, further increasing weapon diversity among species. Overall, our results may help explain how sexually selected weapons have evolved into the diversity of forms seen today.

Keywords: insects, macroevolution, male–male competition, performance, sexual selection, weapons

1. Introduction

Sexual selection has led to the evolution of some of the most bizarre and elaborate traits seen in animals [1], and based on some estimates, almost half (approx. 44%) of these traits are used in contests between males [2]. Contest-related traits include numerous types of weapons, such as the branching antlers of deer, the enlarged claws of fiddler crabs and the elongated horns of rhinoceros beetles [3,4]. These sexually selected weapons may represent up to a third of an individual's body mass (e.g. claws in fiddler crabs [5]) and can be more than twice an individual's body length (e.g. front legs in harlequin beetles [6]). Moreover, these weapons can differ dramatically in their size and shape among closely related species [3]. For example, among species of dung beetles, horns can resemble crowbars, tridents and spears [7]. Yet, the evolutionary mechanisms that underlie the diversification of these weapons among species remain unclear.

Several hypotheses have been proposed to explain the diversity of weapon shapes among species [8]. One of the most well-supported hypotheses proposes that weapon diversification results from differences in how and/or where the weapons are used, called the divergent-context hypothesis [3,9,10]. For example, bovids that wrestle with their opponents are more likely to have horns with long central arches [9]. A similar association between fighting behaviour and weapon shape has been observed in rhinoceros beetles as well [11].

The divergent-context hypothesis is demonstrably important, but another straightforward hypothesis about weapon evolution has yet to be tested. Although weapons can potentially have many functions [3,8], they are primarily used to fight [8]. Thus, a crucial but still untested hypothesis is that weapon diversification is driven by selection for weapon modifications that provide an advantage in combat (i.e. the fighting-advantage hypothesis [8]). There are several ways that weapon modifications could provide a fighting advantage. For weapons that inflict injuries during fights, both theoretical [12–17] and empirical evidence [18,19] suggests that damage can have an important role in determining the contest winner. Thus, one way a fighting advantage could manifest itself is through an increase in the amount of damage that a weapon causes.

Here, we use data from coreid insects (Hemiptera: Coreidae) to test the fighting-advantage hypothesis for the first time. Coreids offer an excellent system to test this hypothesis. First, in contrast with most other systems, the frequency and severity of injury from male–male combat can readily be quantified (figures 1 and 2), as explained below. Moreover, experimental manipulations have demonstrated that this damage can influence fighting outcomes [19]. Second, there is considerable diversity in weapon form across the clade (figure 3), and similar weapons have evolved repeatedly across the group, providing statistical power for comparative analyses [21]. This weapon diversity involves variation in the morphology of their enlarged hind legs (e.g. femur size, number of spines), which are used in male–male competition over access to females and territories [20,22–26]. Despite the diversity of weapon forms, coreid weapons are used in a behaviourally similar manner—to squeeze their opponents [20,22–26].

Figure 1.

Figure 1.

Open in a new tab

Weapons and wing damage in the coreid bug Acanthocephala femorata. Male A. femorata (pictured) use their enlarged, spined hind legs to compete with other males over access to females and resources [20]. In this species, there are prominent spines on the inner surface of the femora. When these competitions escalate, there can be severe damage to the forewings (figure 2). In the individual shown, there are notable punctures on the left forewing. (Online version in colour.)

Figure 2.

Figure 2.

Open in a new tab

The severity of injury from male–male competition in three species of coreid bugs. The density plots show the distribution of damage severity (maximum diameter of a puncture) to the forewings across all measured individuals of (a) Mictis profana (n = 47 punctures), (b) Acanthocephala femorata (n = 254 punctures) and (c) A. declivis (n = 60 punctures). Dashed tan lines represent mean puncture size and solid tan lines represent median puncture size. For reference, the largest puncture wounds in each of the pictured forewings (from left to right) are (a) 0.202, 0.461 and 1.607, (b) 0.088, 0.202 and 1.606, and (c) 0.077, 0.235 and 1.075 mm. All wing pictures were modified (backgrounds removed) to help visualize puncture wounds. A photographed example of each species's hindlimb weapon (in lateral view) is shown above the distributions. (Online version in colour.)

Figure 3.

Figure 3.

Open in a new tab

Weapon diversity, damage severity and phylogeny among 17 species of coreid bugs. Weapon diversity was quantified using a phylogenetic principal component analysis. The first two principal components (PCs) explained 90% of the variation in the data (electronic supplementary material, table S4). PC1 reflects size. PC2 reflects variation in weapon shape. For example, weapons of Thasus neocalifornicus (no. 17) and Acanthocephala femorata (no. 2) have highly negative PC1 scores because they are large (in terms of absolute size). However, T. neocalifornicus has a higher PC2 score than A. femorata because it has a wider femur, fewer prominent spines, and the longest spine is halfway down the tibia. As PC2 increases so does relative damage severity (figure 4). The six selected hind leg images capture most of the observed diversity in weapon size and shape (from species 2, 4, 8, 13, 15, 17 in the tree; the image contacts the corresponding species number). Colour corresponds to relative damage severity, with purple being more severe and tan being less. Numbers on the graph correspond with numbers on the phylogeny. The phylogeny is time-calibrated, and the scale bar indicates the branch length associated with 5 million years. A photographed example of each species weapon is provided directly across from each species tip so that variation in weapon shape can more easily be visualized. (Online version in colour.)

When coreids use their weapons to squeeze their rivals, the rival's wings can become damaged (figures 1 and 2) [23]. Damage in this context is permanent because adult coreids are in their terminal moults and hemimetabolous insects cannot regenerate without moulting [27,28]. This damage also appears costly. The melanization that is often observed around the puncture wounds (figure 2) is indicative of an immune response, which can be metabolically expensive [29,30]. Moreover, wing damage has been shown to reduce flying ability in insects [31,32]. Thus, the wing damage observed here may have implications for an individual's ability to find mates, forage and evade predators. Such costs probably explain why wing damage can influence fighting outcomes in this clade [19].

The fighting-advantage hypothesis leads to clear macroevolutionary predictions in coreids. One way that a weapon modification may provide a fighting advantage is by increasing the amount of damage that the weapon can induce [12–19]. Therefore, this hypothesis would be supported if there is a positive relationship between particular weapon modifications (e.g. additional spines, increased femur width) and the severity of damage inflicted on conspecifics by these weapons. We tested this hypothesis by quantifying weapon diversity and damage from male–male competition among 17 coreid species. We then used phylogenetic generalized least-squares (PGLS) regressions [33] to determine whether weapon modifications (e.g. spines, femur width) were positively associated with damage severity.

2. Methods

(a). Sampling

We selected 17 coreid species to include based on: (i) their presence in a time-calibrated, multi-locus phylogeny for the family; (ii) the presence of male–male competition in these species; and (iii) their representation of the diverse weapon morphologies that occur in coreids. The phylogeny reported by Emberts et al. [21] included 47 coreid species. Among these 47 species, 18 have non-weaponized hind legs (hind legs that resemble their middle legs) and 29 have weaponized hind legs (hind legs that were enlarged and had spines). Among these 29 species, there is published evidence that weaponized hind legs are used in male–male combat in the following species: Acanthocephala declivis [23], Ac. femorata [20], Mictis profana [25], M. longicornis [34], Leptoscelis tricolor [24], Narnia femorata [26] and Leptoglossus gonagra (synonymous with Leptoglossus australis [22]). Moreover, our previous work [21] revealed several other species that had wing damage that was characteristic of male–male competition. Therefore, these species were included as well (Acanthocephala terminalis, Anoplocnemis phasiana, An. curvipes, Phthiacnemia picta, Physomerus grossipes, Piezogaster calcarator and Thasus neocalifornicus). Finally, we wanted to ensure that the diversity of weapon forms across the 47 phylogenetically sampled species was represented. Therefore, we also included Euthochtha galeator, Merocoris typhaeus and Zicca taeniola. Their unique weapon morphologies are apparent in figure 3 (numbers 8–10).

We did not include species with non-weaponized hind legs in our study because doing so could bias the results. Specifically, since species with non-weaponized legs should have no damage from these weapons, we predicted that including species without weapons would only strengthen the observed relationships between damage and weapon morphology. We performed analyses after adding in 11 species with non-weaponized legs and found that the relationship reported in the main text (see below) indeed became stronger (electronic supplementary material, appendix S1). Thus, excluding species with non-weaponized legs is a more conservative approach.

For most species, the specimens used were collected by us and were all sampled from a single locality per species (electronic supplementary material, table S1), to avoid potentially confounding effects of within-species geographic variation on the analyses. However, museum specimens and specimens collected by colleagues [21] were also used to include four additional species. For three of these species (Anoplocnemis phasiana, Physomerus grossipes and Mictis longicornis), most specimens were from the small island of Singapore (total area = 722 km2), although some were from neighbouring countries. Additionally, sampled specimens of Anoplocnemis curvipes were from across southeastern Africa. We used an ANOVA to confirm that these four species had a similar standard deviation in body size as the other 13 species (F1,15 = 0.065, p = 0.802), which were each collected from a single location.

To quantify the damage from male–male competition, we measured an average of 41.1 males per species (mean = 41.1, median = 32, range = 11–104; electronic supplementary material, table S1). We then measured up to 20 males per species to quantify weapon morphology (mean = 15.2, median = 16, range = 5–20; electronic supplementary material, table S1). Importantly, individuals measured for weapon morphology included both specimens that had wing damage and those that did not. We specifically measured a random subset of the specimens' right hind leg weapons. Only right hind legs were measured given high levels of symmetry between right and left legs [35].

(b). Data collection

To quantify weapon form in coreids, we measured seven hind leg weapon components from each specimen when possible. All linear measurements were taken to the nearest micrometre in ImageJ (version 1.46 [36]) from photographed individuals.

We specifically measured: (i) maximum femur width; (ii) femur length; (iii) tibial length; (iv) length of the longest spine; (v) the location of the longest spine; (vi) distance from the longest spine to the fulcrum; and (vii) the number of prominent spines on the right hind leg (electronic supplementary material, figure S1). The femur width measurement captured the widest portion of the femur and did not include spines. If the widest part of the femur had a spine, the measurement only extended to the spine's base. Femur length was the distance between the trochanter–femur joint and the femur–tibia joint. The tibial length was the distance between the femur–tibia joint and the tibia–tarsus joint. Spine length was the distance between the base and the tip of the longest spine. Location of the longest spine was the distance between the middle of the longest spine's base and the trochanter–femur joint (i.e. the fulcrum). Spines that were distal to the trochanter–femur joint (on the tibia) were treated as negative values for location. Spines that were proximal to the trochanter–femur joint (on the femur) were treated as positive values. The absolute values of the spine location measurements were then used as the distance between the fulcrum and the longest spine. To determine the number of prominent spines, the number of spines that were greater than or equal to approximately half the length of the longest spine were counted. Pronotal width, a proxy for body size in this group [21,37], was also measured. This proxy allowed morphological measurements to be size corrected. The final values used for each species were mean measurements among specimens (electronic supplementary material, table S1).

Tibial flags (a tibial dilation found in some coreids) were not considered weapons for two main reasons. First, tibial flags are thought to instead have a role in predator–prey interactions [38]. Second, there can be extreme sexual dimorphism in tibial flag size, but not in the pattern expected for a trait involved in male–male combat. For example, female Acanthocephala femorata have large tibial flags, but tibial flags in males are practically absent [38].

To assess wing damage from male–male competition, we quantified three aspects of damage: (i) the proportion of individuals per species with one or more wing punctures; (ii) the median number of punctures per damaged individual; and (iii) the median size of the punctures in each species. Size measurements were taken to the nearest micrometre in ImageJ from photographed individuals. Each species's median puncture size was calculated using all puncture measurements (electronic supplementary material, tables S1). The median results were prioritized over the mean because visualizations of the damage data often revealed a right-skewed distribution, making the median a better measure of central tendency (figure 2). However, both measures of central tendency produced similar results (see results and electronic supplementary material, tables S2 and S3). Wing punctures were the focus of this study because wing frays and tears may be due to predation or usage [39].

(c). Statistical analyses

We used two complementary approaches to quantify weapon form in coreids. First, we quantified the diversity of weapon shapes with a phylogenetic principal component analysis (pPCA [40]). The pPCA was carried out in R (version 3.6.0 [41]) using the package phytools (version 0.6–60 [42]). The pPCA included all seven weapon components. We then used the scree plot method to determine the number of PCs to retain [43]. We predicted that PC1 would reflect size, and additional PCs would reflect weapon shape.

Second, we analysed each weapon component individually. We size corrected each weapon component by first conducting PGLS regression with the R package caper version 1.0.1 [44]. All PGLS analyses used the lambda value estimated with maximum-likelihood. For each weapon component, we conducted a PGLS analysis including pronotal width (a body size proxy) as the independent variable and each weapon component as the dependent variable. We then extracted the residuals from each of these models. These residuals were then used as the size-corrected shape for each weapon component. Because we tested multiple traits for this second analysis, we applied a sequential Bonferroni correction to each table of these results [45].

We then conducted PGLS analyses to test whether weapon modifications provide an advantage in combat (fighting-advantage hypothesis). One obvious way that weapon modifications may provide an advantage in combat is by increasing the amount of damage induced. Therefore, we investigated how our three damage metrics were related to each of the seven size-corrected weapon components and to composite weapon shape (PC2). We size-corrected maximum puncture width as described for weapon components. Size correcting puncture width was important because a puncture wound with a diameter of 0.2 mm should have greater fitness costs for a smaller individual than a larger one (a higher percentage of the wing missing).

Given that this is one of the first studies to quantify the damage from male–male competition among species, we also investigated whether the frequency of damage observed among species was associated with: (i) absolute weapon size (PC1), (ii) intraspecific variation in femur width or (iii) intraspecific variation in pronotal width. We specifically investigated whether absolute weapon size was associated with the frequency of damage because it has previously been proposed that the most enlarged and elaborated weapons among species rarely inflict damage [3], and our data allowed us to test this hypothesis. Previous studies have also found that coreid fights are more likely to occur when rivals are similar in size [46]. Thus, we wanted to know whether species that had more similar body sizes or femur sizes among sampled individuals (i.e. lower standard deviations for these traits) were more likely to have damage, potentially reflecting an increase in the frequency of damage-inducing fights.

For all comparative analyses, we used a recent time-calibrated phylogeny of Coreidae [21]. This phylogeny was estimated using a maximum-likelihood analysis of 567 loci. The tree was then dated using penalized likelihood [47] and four fossil calibration points. The phylogeny had very strong bootstrap support (all nodes with bootstrap values ≥90%). Therefore, performing comparative analyses across a distribution of trees from these data to address topological uncertainty should yield identical results (but no such distribution was available from these likelihood analyses). This 62-species phylogeny was then pruned to include only the 17 coreid species sampled here. All phylogenies, code, and other data are available on Dryad (see ‘Data accessibility’ below).

3. Results

To quantify the diversity of weapon forms among the sampled coreid species, we conducted a pPCA with seven weapon components (electronic supplementary material, table S1). The scree plot revealed an obvious break in eigenvalues between the second and third PCs (electronic supplementary material, figure S2). Thus, we focused on the first two PCs here. PC1 explained 50.73% of the variance. Most variables related to weapon size loaded strongly onto PC1 (figure 3; electronic supplementary material, table S4). PC2 explained 39.02% of the variance. Femur width, number of prominent spines and two measures of location of the longest spine all loaded strongly on PC2 (electronic supplementary material, table S4). Since PC2 included most variation in weapon morphology that is uncorrelated with size (PC1), we refer to PC2 as weapon shape.

To quantify forewing damage in each species, we measured: (i) the proportion of individuals that had at least one wing puncture; (ii) the median number of punctures per damaged individual; and (iii) the median size of punctures. We observed punctured wings in 13 of the 17 sampled species. The four species without damage were Leptoscelis tricolor, Merocoris typhaeus, Narnia femorata and Zicca taeniola. Thus, none (0%) of the sampled individuals in these species had punctured wings. Our main results are consistent whether or not we included (see below) or excluded (electronic supplementary material, tables S5–S6) these four species from our analyses. For species with one or more punctures observed, the proportion of injured individuals ranged from 5 to 50%. The size of puncture wounds also varied among species. The largest were in Physomerus grossipes (median width = 0.295 mm), and the smallest in Phthiacnemia picta (median width = 0.044 mm; electronic supplementary material, table S1). Physomerus grossipes also had the fewest punctures per damaged individual (median puncture number = 1) and one of the lowest proportions of damaged individuals (12%; electronic supplementary material, table S1), suggesting these two measures may be negatively related with puncture size. However, phylogenetic (PGLS) regressions showed that neither median puncture number (t = 0.806, p = 0.433, r2 = 0.042) nor proportion of injured individuals (t = −0.836, p = 0.416, r2 = 0.045) were significantly related to relative puncture size across all species. However, median puncture number and the proportion of damaged individuals were positively related (t = 3.594, p = 0.003, r2 = 0.463).

The main focus of our study was to determine whether weapon modifications were positively related to damage severity, which would support the fighting-advantage hypothesis. Thus, we first investigated whether weapon shape (PC2) was associated with our three damage measures using PGLS regressions. We found that weapon shape (PC2) was positively related to relative puncture size (t = 2.666, p = 0.018, r2 = 0.322; figure 4). However, weapon shape (PC2) was not significantly related to the median number of punctures per damaged individual (t = 0.080, p = 0.937, r2 < 0.001) nor the proportion of injured individuals (t = −0.716, p = 0.485, r2 = 0.033). Thus, weapon shape was associated with an increase in puncture size, but not the frequency with which punctures occurred. Analysing each weapon component individually revealed comparable results (table 1; electronic supplementary material, tables S7–S8). For example, we found that relative puncture size increased with both femur width and the distance between the longest spine and the fulcrum (figure 4).

Figure 4.

Figure 4.

Open in a new tab

Weapon modifications are positively related to damage severity among coreid species. (a) PC2 (weapon shape metric) increases with relative damage severity (t = 2.666, p = 0.018, r2 = 0.322). Both femur width and the distance between the longest spine and the fulcrum strongly load onto PC2 (electronic supplementary material, table S4). When we analyse femur width (b) and distance from the longest spine to the fulcrum (c) as individual weapon components, both traits increase with relative damage severity (femur width: t = 3.356, p = 0.004, r2 = 0.429; distance from the longest spine to the fulcrum: t = 3.131, p = 0.007, r2 = 0.395). Regression lines are based on PGLS, but the raw size-corrected data are shown here for ease of interpretation. Numbers in data points correspond with numbers in figure 3. (Online version in colour.)

Table 1.

Results of PGLS regressions between weapon components (dependent variables) and median damage severity (independent variable). Italics indicate significant results after applying a sequential Bonferroni correction. Results using mean damage severity are given in electronic supplementary material, table S3.

weapon component t p r2
femur width 3.356 0.004 0.429
femur length −0.095 0.926 0.001
distant from longest spine to fulcrum 3.131 0.007 0.395
spine location −2.058 0.057 0.220
spine length 0.670 0.513 0.029
tibial length 0.332 0.744 0.007
number of prominent spines −3.111 0.007 0.392
Open in a new tab

We also tested whether the amount of intraspecific variation in pronotal width, intraspecific variation in femur width or absolute weapon size were associated with the frequency of damage observed among species. We found that damage frequency was positively associated with intraspecific variation in both pronotal width (t = 2.681, p = 0.017, r2 = 0.324) and femur width (t = 4.670, p < 0.001, r2 = 0.593). Thus, species with adult males that were more dissimilar in size (i.e. higher standard deviations) were more likely to have damage from fighting, the opposite of our prediction. We also found that the proportion of damaged individuals increased as absolute weapon size increased (PC1; t = −3.343, p = 0.004, r2 = 0.427).

4. Discussion

Sexually selected weapons show a bewildering diversity of shapes across many animal groups, but the mechanisms underlying this diversity remain incompletely understood. Here, we tested the fighting-advantage hypothesis [8] as an explanation for this diversity. Based on an analysis of 17 coreid species, we found that certain weapon morphologies (e.g. relatively wide femurs) inflict more severe damage than others. Both theoretical [12–17] and empirical evidence [18,19] suggests that damage can have an important role in determining the contest winner. Thus, our results support the fighting-advantage hypothesis. This hypothesis may explain much of the weapon diversity seen in animals because most sexually selected weapons have a crucial role in fights [3,4,8].

The support for the fighting-advantage hypothesis found here suggests that there is selection for weapon modifications (e.g. a wider femur) that provide an advantage in combat. If there is indeed selection to improve weapon performance, then why have all coreid weapons not evolved to look the same? The diversity of weapon shapes observed in coreids may simply reflect the fact that species are evolving towards a weapon shape that optimizes their performance, and that some species are closer than others to reaching it. Alternatively, conflicting selective pressures may be inhibiting species from reaching their weapon performance optimum, further contributing to the maintenance of weapon diversity in this clade.

Several hypotheses have been proposed to explain weapon diversification [8], and these hypotheses are not mutually exclusive. As a result, multiple hypotheses may help explain weapon diversity within a clade. For example, one reason that Mictis profana may have a narrower femur than Physomerus grossipes is because of selection to reduce the cost of weaponry in M. profana (i.e. divergent cost hypothesis [48]). In coreids, weapon investment comes at the expense of testes investment [49–51]. Therefore, if testes are more important for reproductive success in M. profana than P. grossipes, then selection to reduce the cost of weaponry (to increase investment in testes) could potentially explain why M. profana have narrower femurs. A similar argument could be made for the divergent-context hypothesis (via difference in fighting behaviours). Previous studies on coreid fighting behaviour have all found that coreids use their weapons to squeeze their rivals [20,22–26]. However, in addition to using their weapons to squeeze, at least one coreid species (i.e. Narnia femorata) also uses their weapons to kick their rivals [46]. Thus, species that exclusively rely on their hind leg weapons for squeezing may require relatively wide femurs (with larger squeezing muscles [52,53]; for a biomechanical explanation of the results, see electronic supplementary material, appendix S2). By contrast, species that squeeze and kick (e.g. Narnia femorata) may be able to win some fights with kicks alone (but see [46]). As a result, species that kick and squeeze may have relaxed selection on femur width.

Another way that weapon diversity can arise is if different weapon forms yield similar performance outcomes (i.e. many-to-one mapping of form to function [54–56]). This can be conceptualized through a framework similar to that of an adaptive landscape [57]. In this case, weapon performance is on the vertical axis, while the weapon form is on the horizontal plane (figure 5). If this landscape has multiple performance peaks, then selection on morphological traits to be more effective can drive different clades towards different optima. Simulations have shown that this overall process can increase morphological diversity [54]. In coreids, for example, both Physomerus grossipes and Acanthocephala terminalis inflict severe damage (figure 5). However, their weapon forms differ: P. grossipes has its longest spine on its tibia and A. terminalis has its longest spine on its femur. Thus, coreids can successfully inflict damage onto their opponents' wings if their prominent spine is on either their femur or tibia (i.e. the ability to inflict wing damage is not dependent upon the location of the longest spine). This pattern suggests that the weapon performance landscape in coreids has multiple optima (figure 5). It should be noted that other factors besides weapon morphology (e.g. behaviour) probably contribute to this multi-optima pattern as well.

Figure 5.

Figure 5.

Open in a new tab

A weapon performance landscape for coreids. This figure is a revisualization of figure 3, but in three-dimensional space. The vertical axis shows the amount of damage that the weapon inflicts in each species (relative damage severity: a size-corrected measure of median puncture size), a measure of weapon performance. The horizontal plane is weapon diversity on two PC axes. This visualization of the data shows that there are at least two performance optima (which can also be visualized in a contour plot; electronic supplementary material, figure S3). One performance optimum is associated with species that have their longest spine on the femur (e.g. illustrated here with Acanthocephala femorata, no. 2). The second optimum is associated with species that have their longest spine on the tibia (e.g. Mictis profana, no. 14). Selection on weapons to be more effective should generally drive lineages towards their closest performance peak. The two species shown illustrate that two different weapon morphologies can inflict almost identical amounts of damage. Numbers in data points correspond with numbers in figure 3. (Online version in colour.)

Overall, we found that much of the weapon diversity in coreids may be explained by the presence of multiple performance peaks and by species that have (for various reasons) failed to reach those peaks. Failure to reach a weapon performance peak may simply reflect the fact that a species is evolving towards its performance optima, but has yet to reach it. However, other hypotheses could also be invoked to explain why some species have not reached these peaks. Thus, future work should continue to investigate other hypotheses for weapon diversification in this clade, in addition to the fighting-advantage hypothesis.

Our study also challenges the idea that damage from male–male competition is uncommon in species with extravagant weaponry [3]. For example, in 6 of the 17 sampled weapon-bearing species, more than 20% of measured males had damage (electronic supplementary material, table S1). Moreover, larger weapons were associated with more frequent damage. Our study is not the first to document a high frequency of combat damage in species with elaborate weapons. Intraspecific competition in porcelain crabs [58] and intrasexual competition in Japanese horned beetles [59] also frequently result in substantial injury. In both species, approximately 40% of sampled fighting individuals had fracture wounds [58,59]. At such a high frequency, combat damage could have many additional evolutionary consequences beyond weapon morphology. For example, combat damage might influence the evolution of immune systems and/or promote the evolution of damage-reducing structures [19].

Our results also show that the frequency of damage from male–male competition can vary dramatically, even among closely related species. For example, just within the genus Acanthocephala the frequency of damage among species ranged from 18% to 50% of sampled individuals per species (electronic supplementary material, table S1). Given the potential implications of damage from male–male competition, future work should consider factors that might be responsible for this variation. It is possible that species with higher frequencies of damage have less variation in body size or weapon size and are thus more likely to escalate to high intensity (i.e. damage inducing) interactions. However, we found no evidence to support this hypothesis here, as species with less variation in size (i.e. species that had lower standard deviations) were less likely to have damage, not more. Furthermore, survivor biases might have biased perceptions of the frequency and severity of injury, both in this study and others. We quantified damage by looking at individuals that were collected from the wild and were alive when collected. As a result, we actually quantified the frequency and severity of damage from male–male competition that does not result in death. Death from male–male competition is documented in at least one coreid species (Acanthocephala femorata [20]). This could mean that we underestimated the true frequency and severity of damage. If this is true, the consequences of damage could be even more pronounced. Another potential issue with using wild-caught coreids is that we were unable to age the specimens (beyond being an adult). Thus, we could not verify whether specimen ages were similar across the sampled species. Different distributions of ages for different species could potentially obfuscate the results because older specimens would have had more time to accumulate damage. This could be one reason why we did not find a significant association between weapon shape and the median number of punctures per species. Nevertheless, despite these potential sources of error, we were still able to recover strong relationships between damage severity and weapon morphology.

In summary, we show that much of the weapon diversity in coreids can be explained by how well different weapon morphologies perform at inflicting damage. To our knowledge, this study is the first to show how weapon performance (i.e. at inflicting damage) is related to weapon evolution and diversity among species. We also show that a weapon performance landscape can have multiple performance peaks (figure 5). Thus, selection to win fights could also be driving different lineages towards divergent weapon forms (i.e. many-to-one mapping of form to function), further contributing to overall weapon diversity among species. Finally, in addition to supporting a fundamental hypothesis about the causes of weapon diversity, we show that extravagant weapons can frequently inflict damage (and that larger weapons seem to inflict damage more often). Thus, our results and those of others (e.g. [59]) suggest that damage from male–male combat might be more widespread than previously considered [3].

Supplementary Material

Electronic Supplementary Material
Reviewer comments
rspb20202898_review_history.pdf (974.6KB, pdf)

Acknowledgements

We thank Robert Entzian, Rachel Shepherd and Chelsea Skojec for help with data collection. We thank Christine Miller and Colette St Mary for in-depth discussions about sexually selected weapons. Finally, we thank Cody Howard and three anonymous reviewers for providing helpful comments on earlier versions of the manuscript.

Data accessibility

All the data, phylogenies and code required to recreate our analyses are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.c866t1g5r [60].

Authors' contributions

Z.E. conceived the study, performed the statistical analyses and wrote the first draft of the manuscript. Z.E. and W.S.H collected the data. J.J.W. contributed to study design, statistical analyses and writing. All authors helped edit the manuscript and gave final approval.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the Lee Kong Chian Natural History Museum and the National Science Foundation (DBI-1907051).

References

  • 1.Andersson M 1994. Sexual selection. Princeton, NJ: Princeton University Press. [Google Scholar]
  • 2.Wiens JJ, Tuschoff E. 2020. Songs versus colours versus horns: what explains the diversity of sexually selected traits? Biol. Rev. 95, 847–864. ( 10.1111/brv.12593) [DOI] [PubMed] [Google Scholar]
  • 3.Emlen DJ 2008. The evolution of animal weapons. Annu. Rev. Ecol. Evol. Syst. 39, 387–413. ( 10.1146/annurev.ecolsys.39.110707.173502) [DOI] [Google Scholar]
  • 4.Rico-Guevara A, Hurme KJ. 2019. Intrasexually selected weapons. Biol. Rev. 94, 60–101. ( 10.1111/brv.12436) [DOI] [PubMed] [Google Scholar]
  • 5.Swanson BO, George MN, Anderson SP, Christy JH. 2013. Evolutionary variation in the mechanics of fiddler crab claws. BMC Evol. Biol. 13, 137 ( 10.1186/1471-2148-13-137) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zeh DW, Zeh JA, Tavakilian G. 1992. Sexual selection and sexual dimorphism in the harlequin beetle Acrocinus longimanus. Biotropica 24, 86–96. ( 10.2307/2388476) [DOI] [Google Scholar]
  • 7.Emlen DJ, Marangelo J, Ball B, Cunningham CW. 2005. Diversity in the weapons of sexual selection: horn evolution in the beetle genus Onthophagus (Coleoptera: Scarabaeidae). Evolution 59, 1060–1084. ( 10.1111/j.0014-3820.2005.tb01044.x) [DOI] [PubMed] [Google Scholar]
  • 8.McCullough EL, Miller CW, Emlen DJ. 2016. Why sexually selected weapons are not ornaments. Trends Ecol. Evol. 31, 742–751. ( 10.1016/j.tree.2016.07.004) [DOI] [PubMed] [Google Scholar]
  • 9.Lundrigan B 1996. Morphology of horns and fighting behavior in the family Bovidae. J. Mamm. 77, 462–475. ( 10.2307/1382822) [DOI] [Google Scholar]
  • 10.Caro TM, Graham CM, Stoner CJ, Flores MM. 2003. Correlates of horn and antler shape in bovids and cervids. Behav. Ecol. Sociobiol. 55, 32–41. ( 10.1007/s00265-003-0672-6) [DOI] [Google Scholar]
  • 11.McCullough EL, Tobalske BW, Emlen DJ.. 2014. Structural adaptations to diverse fighting styles in sexually selected weapons. Proc. Natl Acad. Sci. USA 111, 14 484–14 488. ( 10.1073/pnas.1409585111) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Enquist M, Leimar O. 1983. Evolution of fighting behavior: decision rules and assessment of relative strength. J. Theor. Biol. 102, 387–410. ( 10.1016/0022-5193(83)90376-4) [DOI] [Google Scholar]
  • 13.Enquist M, Leimar O, Ljungberg T, Mallner Y, Segerdahl N. 1990. A test of the sequential assessment game: fighting in the cichlid fish Nannacara anomala. Anim. Behav. 40, 1–14. ( 10.1016/S0003-3472(05)80660-8) [DOI] [Google Scholar]
  • 14.Payne RJH 1998. Gradually escalating fights and displays: the cumulative assessment model. Anim. Behav. 56, 651–662. ( 10.1006/anbe.1998.0835) [DOI] [PubMed] [Google Scholar]
  • 15.Parker GA, Rubenstein DI. 1981. Role assessment, reserve strategy, and acquisition of information in asymmetric animal conflicts. Anim. Behav. 29, 221–240. ( 10.1016/S0003-3472(81)80170-4) [DOI] [Google Scholar]
  • 16.Arnott G, Elwood RW. 2009. Assessment of fighting ability in animal contests. Anim. Behav. 77, 991–1004. ( 10.1016/j.anbehav.2009.02.010) [DOI] [Google Scholar]
  • 17.Palaoro AV, Briffa M. 2017. Weaponry and defenses in fighting animals: how allometry can alter predictions from contest theory. Behav. Ecol. 28, 328–336. ( 10.1093/beheco/arw163) [DOI] [Google Scholar]
  • 18.Neat FC, Taylor AC, Huntingford FA. 1998. Proximate costs of fighting in male cichlid fish: the role of injuries and energy metabolism. Anim. Behav. 55, 875–882. ( 10.1006/anbe.1997.0668) [DOI] [PubMed] [Google Scholar]
  • 19.Emberts Z, Wiens JJ. In press. Defensive structures influence fighting outcomes. Funct. Ecol. ( 10.1111/1365-2435.13730) [DOI] [Google Scholar]
  • 20.Mitchell PL 1980. Combat and territorial defense of Acanthocephala femorata (Hemiptera: Coreidae). Ann. Entomol. Soc. Am. 73, 404–408. ( 10.1093/aesa/73.4.404) [DOI] [Google Scholar]
  • 21.Emberts Z, et al. 2020. The evolution of autotomy in leaf-footed bugs. Evolution 74, 897–910. ( 10.1111/evo.13948) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Miyatake T 1993. Male-male aggressive behavior is changed by body size difference in the leaf-footed plant bug, Leptoglossus australis, Fabricius (Heteroptera: Coreidae). J. Ethol. 11, 63–65. ( 10.1007/BF02350007) [DOI] [Google Scholar]
  • 23.Eberhard WG 1998. Sexual behavior of Acanthocephala declivis guatemalana (Hemiptera: Coreidae) and the allometric scaling of their modified hind legs. Ann. Entomol. Soc. Am. 91, 863–871. ( 10.1093/aesa/91.6.863) [DOI] [Google Scholar]
  • 24.Miller CW, Emlen DJ. 2010. Across-and within-population differences in the size and scaling relationship of a sexually selected trait in Leptoscelis tricolor (Hemiptera: Coreidae). Ann. Entomol. Soc. Am. 103, 209–215. ( 10.1603/AN09039) [DOI] [Google Scholar]
  • 25.Tatarnic NJ, Spence JR. 2013. Courtship and mating in the crusader bug, Mictis profana (Fabricius). Aust. J. Entomol. 52, 151–155. ( 10.1111/aen.12011) [DOI] [Google Scholar]
  • 26.Emberts Z, St. Mary CM, Herrington TJ, Miller CW. 2018. Males missing their sexually selected weapon have decreased fighting ability and mating success in a competitive environment. Behav. Ecol. Sociobiol. 72, 81 ( 10.1007/s00265-018-2494-6) [DOI] [Google Scholar]
  • 27.Goss RJ 1969. The principles of regeneration. New York, NY: Academic Press. [Google Scholar]
  • 28.Maginnis TL 2006. The costs of autotomy and regeneration in animals: a review and framework for future research. Behav. Ecol. 17, 857–872. ( 10.1093/beheco/arl010) [DOI] [Google Scholar]
  • 29.Cerenius L, Söderhäll K. 2004. The prophenoloxidase-activating system in invertebrates. Immunol. Rev. 198, 116–126. ( 10.1111/j.0105-2896.2004.00116.x) [DOI] [PubMed] [Google Scholar]
  • 30.Ardia DR, Gantz JE, Strebel S. 2012. Costs of immunity in insects: an induced immune response increases metabolic rate and decreases antimicrobial activity. Funct. Ecol. 26, 732–739. ( 10.1111/j.1365-2435.2012.01989.x) [DOI] [Google Scholar]
  • 31.Combes SA, Crall JD, Mukherjee S. 2010. Dynamics of animal movement in an ecological context: dragonfly wing damage reduces flight performance and predation success. Biol. Lett. 6, 426–429. ( 10.1098/rsbl.2009.0915) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Mountcastle AM, Alexander TM, Switzer CM, Combes SA. 2016. Wing wear reduces bumblebee flight performance in a dynamic obstacle course. Biol. Lett. 12, 20160294 ( 10.1098/rsbl.2016.0294) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Martins EP, Hansen TF. 1997. Phylogenies and the comparative method: a general approach to incorporating phylogenetic information into the analysis of interspecific data. Am. Nat. 149, 646–667. ( 10.1086/286013) [DOI] [Google Scholar]
  • 34.Emberts Z, Miller CW, Li D, Hwang WS, St. Mary CM. 2017. Multiple male morphs in the leaf-footed bug Mictis longicornis (Hemiptera: Coreidae). Entomol. Sci. 20, 396–401. ( 10.1111/ens.12268) [DOI] [Google Scholar]
  • 35.Graham ZA, Kaiser N, Palaoro AV. 2020. Performance, but not size, of hindleg weaponry is sexually dimorphic in the giant mesquite bug (Thasus neocalifornicus). bioRxiv (doi:10.1101/2020.08.03.234385)
  • 36.Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675. ( 10.1038/nmeth.2089) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Procter DS, Moore AJ, Miller CW. 2012. The form of sexual selection arising from male–male competition depends on the presence of females in the social environment. J. Evol. Biol. 25, 803–812. ( 10.1111/j.1420-9101.2012.02485.x) [DOI] [PubMed] [Google Scholar]
  • 38.Emberts Z, St. Mary CM, Miller CW. 2016. Coreidae (Insecta: Hemiptera) limb loss and autotomy. Ann. Entomol. Soc. Am. 109, 678–683. ( 10.1093/aesa/saw037) [DOI] [Google Scholar]
  • 39.Mountcastle AM, Combes SA. 2014. Biomechanical strategies for mitigating collision damage in insect wings: structural design versus embedded elastic materials. J. Exp. Biol. 217, 1108–1115. ( 10.1242/jeb.092916) [DOI] [PubMed] [Google Scholar]
  • 40.Revell LJ 2009. Size-correction and principal components for interspecific comparative studies. Evolution 63, 3258–3268. ( 10.1111/j.1558-5646.2009.00804.x) [DOI] [PubMed] [Google Scholar]
  • 41.R Core Team. 2019. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. [Google Scholar]
  • 42.Revell LJ 2012. phytools: an R package for phylogenetic comparative biology (and other things). Methods Ecol. Evol. 3, 217–223. ( 10.1111/j.2041-210X.2011.00169.x) [DOI] [Google Scholar]
  • 43.Cattell RB 1966. The scree test for the number of factors. Multivar. Behav. Res. 1, 245–276. ( 10.1207/s15327906mbr0102_10) [DOI] [PubMed] [Google Scholar]
  • 44.Orme D, Freckleton R, Thomas G, Petzoldt T, Fritz S. 2013. The caper package: comparative analysis of phylogenetics and evolution in R. R package version 5, 1–36. [Google Scholar]
  • 45.Holm S 1979. A simple sequentially rejective multiple test procedure. Scand. J. Stat. 6, 65–70. [Google Scholar]
  • 46.Nolen ZJ, Allen PE, Miller CW. 2017. Seasonal resource value and male size influence male aggressive interactions in the leaf footed cactus bug, Narnia femorata. Behav. Process. 138, 1–6. ( 10.1016/j.beproc.2017.01.020) [DOI] [PubMed] [Google Scholar]
  • 47.Sanderson MJ 2002. Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach. Mol. Biol. Evol. 19, 101–109. ( 10.1093/oxfordjournals.molbev.a003974) [DOI] [PubMed] [Google Scholar]
  • 48.Emlen DJ 2001. Costs and the diversification of exaggerated animal structures. Science 291, 1534–1536. ( 10.1126/science.1056607) [DOI] [PubMed] [Google Scholar]
  • 49.Somjee U, Allen PE, Miller CW. 2015. Different environments lead to a reversal in the expression of weapons and testes in the heliconia bug, Leptoscelis tricolor (Hemiptera: Coreidae). Biol. J. Linn. Soc. 115, 802–809. ( 10.1111/bij.12544) [DOI] [Google Scholar]
  • 50.Somjee U, Miller CW, Tatarnic NJ, Simmons LW. 2018. Experimental manipulation reveals a trade-off between weapons and testes. J. Evol. Biol. 31, 57–65. ( 10.1111/jeb.13193) [DOI] [PubMed] [Google Scholar]
  • 51.Joseph PN, Emberts Z, Sasson DA, Miller CW. 2018. Males that drop a sexually selected weapon grow larger testes. Evolution 72, 113–122. ( 10.1111/evo.13387) [DOI] [PubMed] [Google Scholar]
  • 52.O'Brien DM, Boisseau RP. 2018. Overcoming mechanical adversity in extreme hindleg weapons. PLoS ONE 13, e0206997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Somjee U, Woods HA, Duell M, Miller CW. 2018. The hidden cost of sexually selected traits: the metabolic expense of maintaining a sexually selected weapon. Proc. R. Soc. B 285, 20181685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Alfaro ME, Bolnick DI, Wainwright PC. 2004. Evolutionary dynamics of complex biomechanical systems: an example using the four-bar mechanism. Evolution 58, 495–503. ( 10.1111/j.0014-3820.2004.tb01673.x) [DOI] [PubMed] [Google Scholar]
  • 55.Alfaro ME, Bolnick DI, Wainwright PC. 2005. Evolutionary consequences of many-to-one mapping of jaw morphology to mechanics in labrid fishes. Am. Nat. 165, E140–E154. ( 10.1086/429564) [DOI] [PubMed] [Google Scholar]
  • 56.Wainwright PC, Alfaro ME, Bolnick DI, Hulsey CD. 2005. Many-to-one mapping of form to function: a general principle in organismal design? Integr. Comp. Biol. 45, 256–262. ( 10.1093/icb/45.2.256) [DOI] [PubMed] [Google Scholar]
  • 57.Simpson GG 1994. Tempo and mode in evolution. New York, NY: Columbia University Press. [Google Scholar]
  • 58.Rypien KL, Palmer AR. 2007. The effect of sex, size and habitat on the incidence of puncture wounds in the claws of the porcelain crab Petrolisthes cinctipes (Anomura: Porcellanidae). J. Crustac. Biol. 27, 59–64. ( 10.1651/S-2728.1) [DOI] [Google Scholar]
  • 59.Siva-Jothy MT 1987. Mate securing tactics and the cost of fighting in the Japanese horned beetle, Allomyrina dichotoma L. (Scarabaeidae). J. Ethol. 5, 165–172. ( 10.1007/BF02349949) [DOI] [Google Scholar]
  • 60.Emberts Z, Hwang WS, Wiens JJ.. 2021. Data from: Weapon performance drives weapon evolution Dryad Digital Repository. ( 10.5061/dryad.c866t1g5r) [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

  1. Emberts Z, Hwang WS, Wiens JJ.. 2021. Data from: Weapon performance drives weapon evolution Dryad Digital Repository. ( 10.5061/dryad.c866t1g5r) [DOI] [PMC free article] [PubMed]

Supplementary Materials

Electronic Supplementary Material
rspb20202898supp1.pdf (1MB, pdf)
Reviewer comments
rspb20202898_review_history.pdf (974.6KB, pdf)

Data Availability Statement

All the data, phylogenies and code required to recreate our analyses are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.c866t1g5r [60].

Articles from Proceedings of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society

RESOURCES