Introduction

Evolutionary biologists have long been captivated by the variation in coloration seen across the animal kingdom and research on coloration has significantly advanced our understanding of evolutionary processes. Inter- and intraspecific signaling are pervasive and have evolved as central elements of key ecological and evolutionary processes. For instance, signaling is essential for species recognition, mate choice, and predator avoidance [1–3]. Intraspecific color divergence has been a central focus of evolutionary research [4–6]. Sexual dichromatism, a form of sexual dimorphism, may occur when selection favors males with exaggerated signals, such as brighter and more colorful morphological traits for mating and male-male competition, whereas selection on females favors dull coloration for background matching [7–12].

Conspicuous communication signals, such as bright and intense coloration, can impose significant costs as they are often used by visual predators to identify prey [13–15]. Therefore it is axiomatic that, compared to females, the risk of detection and predation are high for conspicuous males [16–20]. Nonetheless, evidence suggests that certain adaptations, such as attack deflection, altered behavior, signal partitioning, and communicating privately (i.e. in a manner that is less conspicuous to predators than to conspecifics) can offset the negative consequences of exhibiting vivid signals [14, 21–25]. For instance, Chotard et al. [25] found that the conspicuous hind wing wingtails in the scarce swallowtail butterfly function to deflect a predator’s attack away from vital body parts. Veins in these regions were found to be less resistant to tensile force and break sooner while maintaining fore wing integrity and allowing for an escape. Such adaptations are favored by natural selection as they permit the survival of conspicuous prey [26]. Other processes besides predation risk may also drive the evolution of adaptive morphological traits. For instance, organisms that exhibit sexual selection may undergo adaptation to offset pressures from male-male competition and female mate choice. Male victors from competitions with male rivals benefit by gaining increased mating opportunities [27]; hence, morphological features vital for male-male competition are likely under positive selection pressure in the context of sexual selection.

Wing shape can be an excellent indicator of adaptation to selective pressures such as predator avoidance [25], sexual selection [28], migration [29, 30], and foraging strategies [31]. For instance, wings that are long and narrow are associated with increased speed and agility (the speed at which a turn can be made), whereas wing shapes that are short and wide are associated with reduced speed and increased maneuverability (the radius of the turn that can be made) [32–39]. Specifically, in terms of increased speed, wings that are long and narrow are associated with high acceleration capacity [40]. In addition, body and wing sizes greatly impact the flight performance of flying organisms through changes in wing loading [41]. Wing loading is the measurement of the amount of weight carried by each wing in flight [42].

The order Odonata, composed of dragonflies and damselflies, is one of the oldest lineages of winged insects known for their large size, vivid coloration, conspicuous diurnal behavior, and flight performance [43]. The morphology and shape of dragonfly wings are highly variable [35]; possibly indicating divergent selective pressures on wing shape. The flight performance of a dragonfly can be substantially affected by variation in the shape of fore and hind wings. Johansson et al. [37] found that migratory populations of the globe skimmer dragonfly (Pantala flavescens) had a broader wing base and an overall more slender wing shape compared to non-migratory populations, suggesting that wing shape is important for successful long distance migration. As for wing loading, dragonflies with low wing area relative to body mass exhibit high wing loading which increases flight speed and organisms with high wing area relative to body mass exhibit low loading which increases maneuverability [41]. Organisms with high wing loading exhibit turns with large radius, whereas, organisms with low wing loading exhibit turns with small radius [42].

The common whitetail skimmer dragonfly, Plathemis lydia (Drury, 1773), is sexually dichromatic; males exhibit a conspicuous white abdomen with a median dark band on translucent fore and hind wings, whereas females exhibit a dark brown abdomen with median and terminal dark spots on translucent fore and hind wings (Fig 1). Differences in abdomen and wing coloration between male and female P. lydia are likely attributed to differences in selective pressure. Experiments on clay models suggest that white coloration is more conspicuous to avian predators than duller and darker coloration. For instance, Marshall et al. [44] found that white painted lizard clay models were detected and attacked more often than brown and gray painted lizard clay models. Another study found that the luminance contrast and attack rate were higher on lizard clay models with white markings when compared to lizard clay models with purple markings [45].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Dorsal view of a (a) male common whitetail skimmer dragonfly and (b) female common whitetail skimmer dragonfly (Plathemis lydia).

https://doi.org/10.1371/journal.pone.0303690.g001

Male P. lydia use their white conspicuous abdomen during male-male territorial chases, however, they do not use them during courtship with females [46, 47]. Considering that the visibility of male P. lydia is enhanced, it is likely that they are detected more frequently by predators than females. Predator evasion and male-male territory chases presumably pose the same selective pressures on wing size and shape. For instance, increased acceleration, speed, and agility are essential flying abilities for evading predation [48]. Such flying abilities may allow male P. lydia to accelerate to a maximum speed quickly and evade predation despite increased detection. As for male-male territorial chases, increased speed is associated with effectively herding rival males out of claimed territories [47, 49]; hence, it is likely that increased speed is an essential flying ability for male P. lydia. Considering that females are not under the same pressures as males, it is likely that female P. lydia favor wing shapes that increase maneuverability and gliding instead of speed and agility. Therefore, we hypothesized that male dragonflies will exhibit wing adaptations associated with increased acceleration, speed, and agility to escape from frequent predator attacks and perform male territorial chases, whereas females will exhibit wing adaptations associated with increased maneuverability. We predicted that wing male P. lydia will exhibit narrower and pointier fore and hind wings and higher wing loading compared to females.

Methods

Results

Body length, wing length, area, and wing loading

Residuals for body length, fore and hind wing length, and fore and hind wing area did not deviate from normality (S1 Table). Average and standard deviations for all morphological measurements can be found in S2 Table. We found that male Plathemis lydia dragonflies were significantly longer than female dragonflies (t = -7.4424, df = 33.236, p-value = 1.427e-08) (Fig 2). When running an ANCOVA and using body length as a covariate, we found that male P. lydia dragonflies had significantly shorter fore and hind wings when compared to female dragonflies (Table 1 and Fig 3). In addition, we found that male P. lydia dragonflies had significantly lower fore and hind wing area when compared to female dragonflies (Table 1 and Fig 4). Furthermore, when comparing total wing loading between male and female P. lydia dragonflies, we found that male dragonflies had significantly higher wing loading than female dragonflies (Table 1 and Fig 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Mean body length for male and female common whitetail skimmer dragonfly (Plathemis lydia).

https://doi.org/10.1371/journal.pone.0303690.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3.

Wing length comparisons between male and female common whitetail skimmer dragonfly (Plathemis lydia): (a) Forewing length and (b) Hindwing length. The teal colored circles and slope represent data from female dragonflies whereas royal blue colored triangles and slope represent data from male P. lydia dragonflies.

https://doi.org/10.1371/journal.pone.0303690.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4.

Wing area comparisons between male and female common whitetail skimmer dragonfly (Plathemis lydia): (a) Forewing area and (b) Hindwing area. The teal colored circles and slope represent data from female dragonflies whereas royal blue colored triangles and slope represent data from male P. lydia dragonflies.

https://doi.org/10.1371/journal.pone.0303690.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Wing loading comparisons between male and female common whitetail skimmer dragonfly (Plathemis lydia).

The teal colored circles and slope represent data from female dragonflies whereas royal blue colored triangles and slope represent data from male P. lydia dragonflies.

https://doi.org/10.1371/journal.pone.0303690.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of morphological comparisons using an ANCOVA test.

https://doi.org/10.1371/journal.pone.0303690.t001

Wing shape: Size variation and allometric effect

We found that the fore wing centroid size between male and female P. lydia dragonflies was not significantly different (ANOVA: F = 2.18, df = 1, p-value = 0.1596) (Fig 6). In addition we did not find a significant difference in hind wing centroid size between male and female P. lydia dragonflies (ANOVA: F = 0.08, df = 1, p-value = 0.7762) (Fig 7). When controlling for size, we did not find a significant difference in fore wing shape between male and female dragonflies (Permutation test with 10,000 rounds, p-value = 0.0925) (Fig 6). However, when controlling for wing size, we did find a significant difference in hing wing shape between male and female P. lydia (Permutation test with 10,000 rounds, p -value = 0.0306). We found that male hind wings were narrower at the base and middle with pointier tips than females (Fig 7). For the fore wings, the first principal component accounted for 39.29% of the variance and the second principal component accounted for 25.89% of the variance (S3A Fig). For the hind wings, the first principal component accounted for 56.23% of the variance and the second principal component accounted for 16.70% of the variance (S3B Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6.

(a) Transformation grids and (b) warped outline drawings for fore wing shapes of male and female common whitetail skimmer dragonfly (Plathemis lydia).

https://doi.org/10.1371/journal.pone.0303690.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7.

(a) Transformation grids and (b) warped outline drawings for hind wing shapes of male and female common whitetail skimmer dragonfly (Plathemis lydia).

https://doi.org/10.1371/journal.pone.0303690.g007

Discussion

Our results show that male Plathemis lydia have a significantly longer body length, have significantly shorter fore and hind wings relative to body length, and higher wing loading when compared to females. We also found a significant difference in wing shape between male and female P. lydia, where the hind wings of males were narrower at the base and middle with pointier tips than the wings of females. Fore and hind wing centroid size between male and female P. lydia were not significantly different, showing similar trends to fore and hind wing area when not corrected for body length. Plathemis lydia is a sexually dichromatic organism, where males exhibit a conspicuous white abdomen and females exhibit a dark brown abdomen. The conspicuous abdomen of male P. lydia are not used during courtship with females but instead during male-male territorial chases [46, 47]. Therefore, due to enhanced visibility, it is likely that male P. lydia are detected more frequently by predators than females. Consequently, differences in selective pressures due to sexual dichromatism may have driven the disparities found in the fore and hind wing between male and female P. lydia.

Plathemis lydia exhibits male-biased sexual size dimorphism, where the body length of males are significantly longer than females. Sexual size dimorphism has been widely observed throughout animal taxa including birds [59], anurans [60], fish [61], and insects [62]. Across taxa, the evolution of larger body sizes has been found to be favored by both natural and sexual selection with larger body sizes being associated with higher fitness [63]. Larger body sizes may be sexually advantageous in predator escape [64], resource competition [65], and competition for mates [12].

Sexual selection through male-male competition strongly favors larger male body sizes, contributing to the evolution of male-biased sexual size dimorphism [66]. For instance, territorial bees have undergone selection for larger body sizes compared to females due to their highly territorial behavior and male-male competition as well as the resource defense polygyny mating system of the species [67]. Male-biased sexual size dimorphism has also been associated with an increase in male body size plasticity that is likely driven by sexual selection on males in insect species [68]. In the order Odonata, body size of territorial species is linked to measurements of male fitness such as longevity, mating rate, lifetime mating success, and territorial success [69, 70]. These measurements of fitness are potential drivers of male-biased sexual size dimorphism in P. lydia, especially considering the male-male competition over territorial resources in this species [71].

Male-male territorial chases are energetically costly and having wings that maximize abilities associated with territorial defense is essential [72]. Territorial chases in dragonflies often result in males being evicted from a claimed territory, hence, increased speed is a vital flight performance [47, 49]. Lohmann et al. [73] found, via simulation, that dragonflies with higher speed produce overshooting, which is an adaptive feature of male-male territorial chases. Overshooting is an aggressive interception strategy to pursue and effectively herd rival males from a territory while avoiding collision. Hence, morphological wing features that increase overall speed and energy efficiency should be selected for in dragonflies that perform male-male territorial chases.

Energetically efficient flight is dependent on wing shape, which is tightly linked with flight performance [40, 74]. For instance, wing shapes that reduce flight costs are long and narrow and associated with increased speed and agility [32–39]. Berwaerts et al. [40] found that long and narrow wings are correlated with high acceleration capacity. Male P. lydia have narrower hind wings compared to females suggesting they are adapted for high acceleration during territorial chases. However, P. lydia males have short fore and hind wings relative to body length. One explanation for shorter fore and hind wings is that the decrease in wing area allows for higher wing loading [41]. Wing loading, the ratio of weight to wing surface area of an organism, plays a significant role in flight speed and flight maneuverability [41, 42, 75]. Organisms with high wing loading have a smaller wing area relative to their mass which increases their flight speed whereas organisms with low wing loading have greater wing area relative to their mass which increases their flight maneuverability [41, 42]. It is likely that male P. lydia have an increase in energetically efficient flight due to the shape and size of their hind wings compared to females; which may be vital considering the male-male territorial chases that males perform. Additionally, in territorial odonate species, males with longer wing lengths have been found to defend territories for fewer days compared to males with shorter wings, suggesting that this may be another advantage of exhibiting shorter wings during male-male competition [76]. Increased agility and speed due to narrow wings and higher wing loading is likely to result in more effective exclusion of competitors from a dragonflies’ territory.

In addition to male-male territorial related stressors, male P. lydia are likely to experience predator related stressors due to enhanced visibility. Organisms with bright conspicuous coloration are more likely to be detected by predators than organisms with dull coloration, especially males in sexually dichromatic species [13–18]. Increased speed and agility are essential flying abilities for evading predation [48] and, similarly to male-male competition, morphological features that increase these flight performances should be selected for. Considering that male P. lydia exhibit enhanced visibility, it is likely that they are detected more often than females and therefore, more susceptible to predation. The morphological features that may have been adapted for territorial chases, narrow wings and high wing loading, may also be adapted for the speed and agility required for predator evasion.

Compared to male P. lydia, females have longer fore and hind wings relative to body length and lower wing loading. In addition, we found that females exhibited hind wings that were wider at the base and middle with blunt tips. Organisms that exhibit wide wings and lower wing loading are likely to have increased energy demand with reduced flight speed and increased flight maneuverability [32–39, 41]. Dakin et al. [77] found that species with low wing loading display turns that have faster rotations and are sharper. As a result, speed may not be an important aspect of flight performance for female P. lydia and instead high maneuverability has been selected for. Maneuverability in flight has two primary components, turning radius and speed, that may be subject to different selective pressures [75]. Components of maneuverability are impacted by the allocation of mass, specifically the center of mass for a species [75, 78]. Mass allocation could explain differences between male and female wing areas relative to body length. High maneuverability allows organisms to evade predation as it is associated with the ability to complete turns with small radii [42, 52]. Prey are generally smaller than their predators and this allows prey species to make turns that are quicker and sharper than predators, allowing them to escape predation despite being slow flyers [79]. In addition, studies suggest that the basal lobe of the hind wing is associated with gliding [37, 80] and therefore, it is likely that female P. lydia have increased gliding performance compared to males due to exhibiting wider hind wings.

Conclusions

In conclusion, we investigated wing morphometric differences between sexes of a sexually dichromatic dragonfly, P. lydia). Male P. lydia display a conspicuous white abdomen whereas females display a dark brown abdomen. Due to differences in conspicuousness, it is likely that males are detected more often than females, and therefore, experience an increased level of predator related stressors. In addition, male P. lydia exhibit male-male territorial chases, which may cause increased sexual selection pressures. We found that male P. lydia had significantly shorter fore and hind wings relative to body length and higher wing loading when compared to females. In addition, we found that male P. lydia had hind wings that were narrower and pointier than females. The shape and size of male wings suggest that there have been adaptations associated with increased flight speed, agility, and acceleration capacity. Female P. lydia have wider hind wings that are likely adapted to increase flight maneuverability in the form of smaller turning radii. Considering that sexes of P. lydia dragonfly experience different selective pressures, the sexes have likely adapted different wing morphologies that offset these pressures.This is the first study to demonstrate wing morphological differences between the sexes in a sexually dichromatic dragonfly where males and females are undergoing different selective pressures due to variation in conspicuousness.

Supporting information

Acknowledgments

We are grateful to Michael Reynolds and Brandon McNeil for their assistance in wrangling dragonflies and Dr. Michael Brewer for allowing us to use his insect nets. In addition, we are grateful to Dr. Michael Brewer and Dr. Mysia Dye for comments on this manuscript. Lastly, we are also grateful to the anonymous reviewers for their insightful comments and suggestions that improved this manuscript.