Testing bird-driven diurnal trade-offs of the moon moth's anti-bat tail

Abstract

Traits are often caught in a dynamic tension of countervailing evolutionary pressures. Trade-offs can be imposed by predators evolutionarily curtailing the conspicuousness of a sexually selected trait, or acting in opposition to another natural selection pressure, for instance, a different predator with a divergent hunting strategy. Some moon moths (Saturniidae) have long hindwing tails that thwart echolocating bat attacks at night, allowing the moth to escape. These long tails may come at a cost, however, if they make the moth's roosting form more conspicuous to visually foraging predators during the day. To test this potential trade-off, we offered wild-caught Carolina wrens (Thryothorus ludovicianus) pastry dough models with real Actias luna wings that were either intact or had tails experimentally removed. We video recorded wrens foraging on models and found that moth models with tails did not experience increased detection and attack by birds. Thus, this elaborate trait, while obvious to human observers, does not seem to come at a cost of increased avian predator attention. The evolution of long hindwing tails, likely driven by echolocating predators at night, does not seem to be limited by opposing diurnal constraints. This study demonstrates the importance of testing presumed trade-offs and provides hypotheses for future testing.

Keywords: Carolina wren; Saturniidae; elaborate trait; evolutionary trade-offs; predator–prey.

Figure 1.

Hindwing tails in Actias luna moths do not come with a trade-off against foraging birds. (a) We pit Carolina wrens (Thryothorus ludovicianus) in outdoor flight cages against two different moth pastry dough models, using real A. luna wings. The treatments were either natural wings (tailed) or wings with the tails cut-off (tailless). (b) We presented these models on branches of sweetgum, A. luna's preferred host, and measured which model the bird attacked first. (c) Birds did not preferentially strike either treatment first. Thus, tails do not seem to draw avian predator attention, nor do they afford roosting moths safety by breaking search image. Central points depict mean marginal effects and error bars are 95% confidence intervals.