Skip to main content
Sign In

University of Colorado Denver College of Liberal Arts and Sciences

Biology Logo

The Swallow Lab


Sexual selection is responsible for much of the morphological diversity among animals and is often expressed in the form of male ornamental displays. Sexual selection often results in exaggeration of morphological and behavioral characters when individuals of a given sex, usually males, experience a mating advantage through the mechanisms of female choice or male competition.  Almost all recent theoretical models assume that elaborate male secondary sexual characteristics are costly to produce and to maintain; thus, the mating benefit of such traits (sexual selection) is offset by other fitness costs (natural selection).  

Stalk-eyed flies (Diptera; Diopsidea) are emerging as an important model system for understanding how choosy females drive the evolution of showy male traits.  All flies in the family Diopsidae are characterized by the elongation of the head into long stalks, with eye bulbs and antennae laterally displaced on these stalks.  In some species, this elongation is so extreme that eye span exceeds body length.  Eyestalks are an important feature in diopsid mating systems; in sexually dimorphic species of stalk-eyed flies male mating success is positively correlated with eye span.  Females show a preference for males with longer eye spans.  In addition, males compete for access to and control of prime breeding sites (dangling rootlets under overhanging embankments) and for groups of females by comparing eye span and sometimes, if males are similar in size, in pitched battles.  Such observations indicate that eye span is under current sexual selection.  Furthermore, eyestalk length is heritable and, thus, capable of responding to selection.  A critical gap in this body of work is information regarding constraints on the system.  In order to fully understand population differentiation and what prevents “run-away” elaboration of male eye stalk length, it is important to understand the costs associated with producing and maintaining eyestalks.  

Stalk-eyed flies present a particularly exciting model for investigating trade-offs between sexually exaggerated traits and potentially related physiological and locomotor behavior traits (flight performance).         

Physiological constraints generated by the elongation of eye stalks could create selectively important variation in flight performance and thereby limit further elaboration of the trait.   Selectively important effects could be manifested directly as an increased energetic cost of flying and/or reduced flight performance, or by compensatory changes such as an increase in the relative amount of flight muscle.  The specific questions that my lab is currently trying to address are as follows:
1) Do exaggerated eye stalks result in functional locomotor and energetic costs?
2) Do morphology and locomotor performance show patterns of correlated evolution within the family Diopsidae?
3) Do stalk-eyed flies show correlated evolutionary change in behavior and body design such that locomotor costs of ornaments are reduced?

Studies on flight mechanics suggest that males, especially with larger eye span, may have impaired flight abilities when compared to females (Swallow et al. 2000).  To address question 1,we tested whether the long eye stalks have a negative effect on free-flight and aerial turning behavior by comparing the morphology and the free-flight trajectory of dimorphic male and female Teleopsis dalmanni. At flight posture, when the body is inclined at approximately 70º to the horizontal, the MOI for a rotation about a vertical axis was calculated to be 1.47 fold higher in males than in females. Dimorphism in eye-stalk length is also associated with a 5% increase in male wing length. During free flight both sexes performed aerial turns in the horizontal plane. We found only weak evidence that females turn at a larger angle than males (54 ±31.4 º vs. 49 ± 36.2 º, t-test, p<0.033) and fly faster while turning (9.4 ±5.45 vs. 8.4 ± 6.17 cm s-1, ANOVA, p<0.021). However turning performance of both sexes overlapped, and at the upper 25% of observed turn performance, for a given turn size and flight speed, turn rate in males even marginally exceeded turn rate in females (733 ± 235.3 vs. 685 ± 282.6 ºs-1, ANCOVA, p<0.047). We suggest that the increase in eye span does result in an increase in the mechanical requirements for aerial turning but that male T. dalmanni are capable of compensating for the constraint of longer eye stalks during the range of turns observed. The compensation mechanism is not clear but may involve wingbeat kinematics combined with the observed difference in wing size. 

Furthermore, we studied whether the 2-fold higher moment-of-inertia (MOI) of the male head results in a reduced head rotation velocity during turning (Ribak et al., 2007). We analyzed movies of flies performing walking turns and compared the head kinematics between the sexes. The significance of head rotation to turning was evaluated from turning kinematics of flies with immobilized (glued) heads. Male and female T. dalmanni rotated their heads relative to the surrounding 1.55 (♂) and 1.65 (♀) fold faster than the angular velocity of the body by performing rapid head saccades. During the larger turns flies with immobilized heads were unable to reorient gaze as fast as the control flies. Despite the larger MOI of the head, male T. dalmanni match the head saccade of females suggesting that eye-span elongation is coupled by an adaptation of the neck musculature to rotate the wider head.

Although morphological compensation may minimize some of the costs associated with eye stalks in flight (Ribak and Swallow 2007, Ribak et al. 2009), even slight locomotor impairment may have an effect on Darwinian fitness through predation. To address question number 2, we quantified the anti-predator behaviors and predation risk of male and female Teleposis dalmanni were quantified during pairwise interactions between flies and an actively foraging, generalist arachnid predator (Phidippus audax).  Male and female flies were compared under the assumption that female stalk length is closer to the optimum set by natural selection.  There were significant differences in the behaviors of the sexes, with males spending more time engaged in aggressive actions.  Interestingly, males also exhibited increased survival relative to females.  Within males, survivors did not differ from non-survivors in any of the measures of morphology, including eye span, but did show a significant difference in abdomen bobbing, grooming, and flight. These results provide evidence that behavior, more so than morphology, may be especially important for stalk-eyed flies to effectively elude predators. To test the generality of these results, we are performing similar experiments in larger mesocosms.

To address question 3, we examined whether an increase in eye-stalk length results in an adjustment of wing size and shape to deal with the burden of bearing an exaggerated ‘ornament’ (Ribak et al., 2009). We compared wing morphology among 10 species of stalk-eyed flies that differ in eye-span and degree of sexual dimorphism. Mass-specific wing length differed between males and females in 7 out of the 10 species. Non-dimensional wing shape parameters differed between the species (p<0.001) but mostly did not differ between males and females of the same species. Dimorphism in eye-span closely correlated with dimorphism in wing length (r=0.89, p<0.001) and the correlation remained significant (r=0.81, p=0.006) after correcting for phylogenetic relationships. Once corrected for phylogenetic relatedness, mass-specific wing length of males (but not females) was weakly correlated with mass-specific eye-span (r=0.66, p=0.042). We propose that the observed proportional increase in wing length associated with increased eye-span can facilitate aerial maneuverability which would otherwise be handicapped by the elevated moment of inertia imposed by  the wider head. 

Much of the research described in these web pages is supported in part by NSF CAREER grant (Performance and Fitness Consequences of Insect Ornaments) IOB-0448060 and supplements